Lightning doesn´t solve the scaling problem

[Antihero]

Preface:

I've been thinking about this for a few months now (and please correct me if I'm wrong), but Lightning is absolutely bad for Bitcoin in my eyes.

I'm a person who always thinks far into the future, and doesn't like to discard things after 5 years because they turned out to be shit, so I want to raise my concerns here early to maybe spur one or the other not to rest on half-finished, possibly later useless soluions. Because as we know - innovation must not be slowed down by a "possible solution", we definitely have to continue working on a meaningful scaling for BTC.

Why?

First of all, why do we all love Bitcoin? Because it's brilliant, and totally secure!

Now let's think around the corner a bit. What if Bitcoin is really established in the world? And let us not talk about established as a sort of money / payment solution, but "only" as digital gold. Because the first one would be most worse.
 
Bitcoin is currently able to process around 600,000 transactions daily (this is a roughly rounded number. At a TPS average of 7), the actual number is slightly higher. Since I'm from Germany, I'll take this country as a reference with its nearly 80 million inhabitants.

In 2019, there were about 10.65 million children under the age of 14 living in the Federal Republic of Germany, which means I take them out of financial transactions (although they might also like to have a BTC savings account). Remains round 70 million people, If only 50% of them use bitcoin as a "Save Heaven", which is for me more than logical with a mass-adoption this technology definitely deserves, we are at 35 million people in the small country of Germany alone hold their reserves in Bitcoin. (in the rest of the world there are still 7.7 billion people!!).

It would take the equivalent of 58 days for each of those 35 million people to make a transaction with the mainlayer. If only ONE THIRD of the worldwide humans ever wanted to make a transaction with the mainlayer, this would take 4290 days! As an example, a person who pays into his pension every month could only do so every 4290 days! (earliest!)
 
So far to the problem, this is known to us, and why "we" research things like Lightning.

Unfortunately, we now realize that based on the above figures, not even so-called settlement transactions or individual transactions on my Mainnet wallet would be possible, unless I pay fees that are so high that they are no longer in any relation. EVERYTHING that will ever happen on BTC would only happen on the semi-secure second layer which is still in its infancy after 5 years. (Including all of my savings, which I may have set aside for my retirement, would exist for my lifetime only as LBTC, and possibly be managed by a custodial service).

Where is the advantage to the current system? We trade security for speed and scale (what many criticize about the so-called "altcoins" becomes mandatory for BTC).
Yes, there remains the security of the hard cap that protects us against inflation, but is this worth the insecurity of every individual who may have to lock up their life savings in a smart contract for 50 years because they cannot make a mainnet transaction? In such a case, will the main layer even be able to reliably map the values of LN?

How will the onboarding process look like? If a new person wants to buy BTC, they can't even send the BTC from the exchange to their own wallet because the mempool is always on his limit. Should exchanges also trade LBTC only in the future? In the long term, this will completely eliminate the "good" BTC for LBTC.

I believe that solutions have to scale the main layer at least a little more, we cannot sit back until the supposed hope carrier Lightning is "ready".
What would happen if bitcoin adoption comes in 2021? We would not be ready for it. Neither the Mainlayer, nor LN or Liquid would withstand such an amout of people and their needs.

REMEMBER: this is only the case if we have a truly adopted BTC! (Which most Bitcoiners aiming for ASAP.)

Please don´t give all your hopes only to Second layer solutions.

Let the technical discussions begin.

 

[topcoin360]

You're damn right but there is even a bigger problem. As the LN gets more popular there will be more and more fraudulent transactions using space on the main block chain and with a 1MB block size limit that could be problematic. BTC's creator talked about increasing the block size limit as the demand for space increases but he never talked about segwit or a L2 solution.. in my opinion BCH is the real Bitcoin and ETH is the future! 

[mikeywith]

So we raise the block size to 8mb and then what do we do whenever that isn't enough? I am not against reducing or limiting the scaling problem by increasing block size but it doesn't really solve anything, no matter what you do, you can never beat the ordinary system in terms of speed and cost, blockchain is slow and expensive, this isn't exactly an issue it's more like some of its main characteristics, Paypal payments that are done on one server or a few at most will always outperform blockchain at its main layer, this isn't my opinion it's just how things are, saving a transaction on tens of thousands of computers isn't suppose to be faster and cheaper than visa or PayPal.


The second point which I feel the need to explain since you mentioned that you are from Germany which also goes to folks in the U.S and the other first-world countries, the internet is decades behind in most other places, where I live even with the current block size only a few people can run a full node or directly mine to their own node, 1MB of bandwidth is a big deal here, it's hard to believe because you just pay a fraction of your salary for a 100mbps or even 500mbps connection with unlimited quota, but elsewhere, you can hardly get a 1mbps line with very limited monthly quota, as it stands now, I can't even propagate a 1MB block in a timely manner. By increasing the block size you simply send many players out of the game.

Another point is the fact that your transaction will be saved on tens of thousands of computers and it would take years and billions of dollars to re-do the blockchain, it really has to come at a cost, getting your transaction on the blockchain is more secured than saving your money in the safety deposit boxes at your local bank, it is worth a lot, and someone has got to pay for it,  and at the end of the day, why would you want your 1$ transaction to be saved on the main layer?

One last thing, let's assume we solved the issue of block size, what about confirmation time? say the transaction fee is zero due to unlimited block size, will you be able to pay for coffee when relying only on the main layer? will the shop wait for 10 or 20 minutes until the transaction is confrimed before serving you coffee?

[DaveF]

...I can't even propagate a 1MB block in a timely manner. By increasing the block size you simply send many players out of the game....

At this point in time how many people who do not have good fast internet are really mining to a local node?
Coming from a 1st world county as you put it, I could be seeing this wrong but...
If you are running a mining farm that has enough power to realistically find a block, between the hardware and the cooling and everything else is good bandwidth that much more? I am not being snarky, I just don't know.

If you are running a small home miner as a "lottery" miner. You really should be mining to a solo pool out there. Between hardware issues, network issues, etc. It's probably not worth doing it yourself. Hell, I can barely keep a small server up for 2 minor scrypt coins. And it's on a real (if old) Dell server in a real data center.

-Dave

[FinneysTrueVision]

Lightning isn't meant to solve Bitcoin's scaling issues and neither are sidechains. They are only protocols that extend Bitcoin's functionality while maintaining the base layer secure and decentralized. If we throw that out the window in favor of putting a band-aid on the problem then Bitcoin is no different than any altcoin.

[Wind_FURY]

So we raise the block size to 8mb and then what do we do whenever that isn't enough? I am not against reducing or limiting the scaling problem by increasing block size but it doesn't really solve anything, no matter what you do, you can never beat the ordinary system in terms of speed and cost, blockchain is slow and expensive, this isn't exactly an issue it's more like some of its main characteristics, Paypal payments that are done on one server or a few at most will always outperform blockchain at its main layer, this isn't my opinion it's just how things are, saving a transaction on tens of thousands of computers isn't suppose to be faster and cheaper than visa or PayPal.


The second point which I feel the need to explain since you mentioned that you are from Germany which also goes to folks in the U.S and the other first-world countries, the internet is decades behind in most other places, where I live even with the current block size only a few people can run a full node or directly mine to their own node, 1MB of bandwidth is a big deal here, it's hard to believe because you just pay a fraction of your salary for a 100mbps or even 500mbps connection with unlimited quota, but elsewhere, you can hardly get a 1mbps line with very limited monthly quota, as it stands now, I can't even propagate a 1MB block in a timely manner. By increasing the block size you simply send many players out of the game.

Another point is the fact that your transaction will be saved on tens of thousands of computers and it would take years and billions of dollars to re-do the blockchain, it really has to come at a cost, getting your transaction on the blockchain is more secured than saving your money in the safety deposit boxes at your local bank, it is worth a lot, and someone has got to pay for it,  and at the end of the day, why would you want your 1$ transaction to be saved on the main layer?

One last thing, let's assume we solved the issue of block size, what about confirmation time? say the transaction fee is zero due to unlimited block size, will you be able to pay for coffee when relying only on the main layer? will the shop wait for 10 or 20 minutes until the transaction is confrimed before serving you coffee?

Plus the big blockers have lost the idea of Bitcoin's main value proposition, not speed, but Censorship-resistance. Which requires the network to scale out to be more secure. And it is better to go for more security than less security.

[Antihero]

You guys got me wrong. I´m NOT talking about simply increasing the block size, this would not be work in the long therm. If its not the bandwith who limit this, the disc space will do so definitelly earlyer or later. I also didnt want a discussion between big and small blockers here. Its a fundamental problem a lot of people misunderstand and need to be solved in a rational way, ahead of everyones personal financial interests here!

Lightning isn't meant to solve Bitcoin's scaling issues and neither are sidechains. They are only protocols that extend Bitcoin's functionality while maintaining the base layer secure and decentralized.

you absolutelly missed my point. You can´t even use the base layer as you can read in my first post! THIS IS SOMETHING NEED TO BE SOLVED! simply to say this has to be limited, is in the worst case a death for bitcoin at a mass adoption!

and again: to solve the problem isn´t simply increasing a blocksize or anything, its a complex dilemma that needs a lot of research!

think logically!
 

[ABCbits]

People who learn technical side of Bitcoin (or P2P system in general) already know there's no single solution for scaling problem, even LN have 2 main disadvantage,
1. LN is only meant for micro-transaction
2. LN only useful if you make transaction regularly (at least more than 2 transaction within channel duration)

As for solution on base-layer, there are few option,
1. Schnorr signature
2. Using different number representation on transaction/block (e.g. Increase block size (5%) by using CompactSize instead of UInt)
3. Just increase block size based on lowest growth of computer hardware and internet speed on 3rd/developing world country.

but elsewhere, you can hardly get a 1mbps line with very limited monthly quota

It's naive question, but how limited is the monthly quota? 100GB? 10GB? or just 1GB?

I can't even propagate a 1MB block in a timely manner.

Actual block propagation uses compact block (where it's size only few KB) though

[NotATether]

...I can't even propagate a 1MB block in a timely manner. By increasing the block size you simply send many players out of the game....

At this point in time how many people who do not have good fast internet are really mining to a local node?
Coming from a 1st world county as you put it, I could be seeing this wrong but...
If you are running a mining farm that has enough power to realistically find a block, between the hardware and the cooling and everything else is good bandwidth that much more? I am not being snarky, I just don't know.

If you are running a small home miner as a "lottery" miner. You really should be mining to a solo pool out there. Between hardware issues, network issues, etc. It's probably not worth doing it yourself. Hell, I can barely keep a small server up for 2 minor scrypt coins. And it's on a real (if old) Dell server in a real data center.

-Dave

Most of us here are on CGNAT networks and that means we don't get the privilege of getting static IP addresses unless we buy an expensive fiber line which itself isn't as fast as you think, they can transmit data across the internet at ~50mbps instead of gigabit speeds that would be expected from fiber. And fiber is really only available st universities and large institutions/companies.

The rest of us chaps are stuck with either buying ADSL broadband or a wireless 3G or 4G/LTE data plan. The wireless plans as you'd expect from a third-world country have high latency (because all the remote servers are on some other continent) and have speeds that aren't really suitable for mining or running a full node. We're talking around 10mbps here which is barely enough to push a 1MB block out in a second.

The ADSL lines in my country of residence are actually quite bad and do at most 512kbps up and 4mbps down, but I expect some other third-world countries to have slightly better speeds than this. But even those are still around the speed you get from LTE so you still have all of those lagging connection problems as you do with a data plan. And this is to say nothing about how you are supposed to broadcast your node when your IP is dynamic.

Power isn't a problem though, private generators are readily available here and it isn't too difficult to hire electricians to set up a building's electrical systems for one.

[DaveF]

Replying to all without quoting:

Even here in the US (NY) you don't get static IP unless you pay for it usually along with other "required" upgrades.
Most providers classify it as a business class service which is a lot more. But in reality you don't need static for anything BTC related.

But, outside of the initial download you really don't need that much up or down.

As ETFbitcoin said with compact blocks it's actually a small propagation. You still have to have a copy of the mempool on your node so that is where the bandwidth is needed. \
Block size really does not matter at that point (much) 1M, 2M, 4M if you are running a node you still have to have the full current mempool if you are really mining.

For some reason people have gotten it into their heads that running a node requires a stupid amount of power all the time.
Yeah, for the initial sync more ram more CPU is better but a RPi4 does a fine job of it.

Using Amazon US store:
RPi4 with metal case for cooling and power supply $85: https://amzn.to/3hyi2H2
1TB USB SSD $100: https://amzn.to/3hyznzv

All you need after that is a micoSD card so for under $200 you have a fulling running node.

And here is the fun kicker. You can be running you own lightning node too.
There are at least 2 "nodes in a box" that you can install that do all the work for you.

mynodebtc: https://www.mynodebtc.com/ bitcoin / lightning and some other stuff all ready to go. You can pay $99 to unlock more features if you want but its 100% functional without doing that. AND it allows you to use both your BTC & Lightning node on TOR out of the box.

raspiblitz: https://github.com/rootzoll/raspiblitz A bit better setup, not as user friendly, does a bit less BUT I have had one up close to a year now with no issues that I did not create myself

Not I know in many parts of the world where $200 is a ton of money. I don't have a good answer for that. Yes, you can probably put together a used PC system for a lot less, but other then that I have no good idea.

-Dave

[NeuroticFish]

Let the technical discussions begin.

I will start with criticizing some of your examples. And it won't be that much technical, sorry.
If your BTC savings will be handled by a custodian service, then it will be able to optimize the transactions, making some of them off-chain or grouping (many) more of them in one transaction.

From my understanding on LN, settlements should not be necessary after some time, since the transactions would flow in both directions between nodes, and the nodes that needs settlements often (or at all) will become unproductive and will be shutdown.

Since Bitcoin is meant to become at some point a coin for every day use, LN (or something like that) is necessary in order to avoid the need for waiting at the grocery store many long minutes for blockchain confirmations.
So it has its own use, not necessarily for your problem.

For your problem I think that "the world" will auto-adjust. Your numbers (I'm in criticizing mode, remember? ) forget that if that many people will have Bitcoin, most probably the amounts owned by the most would be under network dust amount, hence cannot be transacted on-chain (and no, it's not simply 21M/3B)

All in all:
1. normal people will use a few Satoshi at the grocery store on LN or similar
2. normal people will probably use custodian service for Bitcoin-as-digital-gold, hence there the number of transaction I expect to be far less than what you predict
3. the few lucky ones with coins in their own wallet will be rich enough to not mind the high network fees for fast processing in case of occasional spend

I think that this kind of problems have to be cleared up before doing research in the direction of changing Bitcoin at its core.

[GGUL]

Plus the big blockers have lost the idea of Bitcoin's main value proposition, not speed, but Censorship-resistance.

Increasing the block has nothing to do with resistance to censorship. This is another misconception that occurs in the Bitcoin community.

In order to introduce censorship in the Bitcoin system, it is necessary to introduce censorship for all mining pools. But the resistance of mining pools to censorship does not depend on the block size.

And it is better to go for more security than less security.

If the main goal is to increase security, and the rest does not matter, then there is a great way to achieve this. Let's make the block size equal to zero. Then all bitcoins will never be lost.

[mikeywith]

It's naive question, but how limited is the monthly quota? 100GB? 10GB? or just 1GB?

It depends on what you can afford, but I'd say the average packages that the average can buy would be anywhere between 20GB to 30GB monthly quota and at a download speed of 1-2 Mbps and an upload of 0.2-0.4 Mbps.

Due to the nature of my job I can afford a 5Mbps up and 1Mbps down with above 100GB of quota and free download at certain periods, but I pay a little bit more than 20% of the average salary for a nurse, school teacher, or a police officers, but eventually when this package doesn't make economical sense (if I lose my job ), I will have to go back to those limited and slow packages.

And if you think that is bad, I have seen worse elsewhere.

All you need after that is a micoSD card so for under $200 you have a fulling running node.

There are at least a few countries where the average wage is below $50 a month, there are a few tens of other countries with monthly wages of below $200 and perhaps only 50 countries or so with wages of above $500 a month, so $200 is a lot of money for the majority of people given that most populated countries have low wages, it has been a while since I looked at the numbers but I can safely bet that at least 8 out of 10 human beings on this planet earth have wages that are below $300.

This isn't exactly bitcoin-related, but the majority of wealth sits in the hands of a minority of other people who are mainly located in EU and USA, it is just the way it is, and for those who can afford a full node with 5-10 hours of a moderate job, I do understand if you find it difficult to believe that others have to work for 2-3 months to afford an SSD that can accumulate the current blockchain size, and as hard as it is now, it will only get worse when the blocksize is increased.

With that being said, not all "poor" people will want to run a bitcoin node since even the current 300GB and the 1MB size is way too huge for them, but when we make suggestions of increasing block size to 16MB or lift the cap altogether, we have to evaluate the technology capabilities worldwide and not just within 5-10 developed countries, just my 2 bytes on the subject.

[GGUL]

As for solution on base-layer, there are few option,
1. Schnorr signature
2. Using different number representation on transaction/block (e.g. Increase block size (5%) by using CompactSize instead of UInt)

Schnorr signatures will give a theoretical 20%, and in practice 10% increase, block optimization will give 5%. Together, this will give a 25-15% increase. To the existing 10 million users per month, this will add 1.5 -2.5 million. It's a drop in the bucket . I wouldn't use the word "scaling" for this.

3. Just increase block size based on lowest growth of computer hardware and internet speed on 3rd/developing world country.

For scaling Bitcoin at the moment, there is no alternative but to increase the block.
Increasing the block by 2 times will give 100%, and increasing it by 10 times will give 1000% scaling. At the same time, this does not require any special sacrifices for full-node holders.
According to my calculations, it is enough for me to buy a 2TB hard drive to easily keep a full node with an enlarged block for 5 years. In fact, a increased by 10 times block and a fees worth cents will be more profitable for me than a 1MB block and a fees of 5-10$ for each transaction.

I'm not even sure whether to focus on the 3rd countries. There's something wrong with that. Let's assume that in developed countries we now have 1 billion potential users, of which 10 million can hold a full node with a 10MB block. That is, increasing the block by 10 times can increase the number of users up to 100 million.
But we look at the 3rd countries and decide not to increase, because they will not be able to.. And what it will give us. As there were 10 million users, so there will remain 10 million. Question: What do we get from not increasing it? Who did we do better for?

p/s/ In order to understand that LN will not be able to scale Bitcoin even to 50 million users, it is enough to be able to use a calculator.
And then there are amazing things:
1) The Core team claims otherwise.
2) The Bitcoin community mostly believes them.

What, both of them do not know how to use a calculator?

[DaveF]

With that being said, not all "poor" people will want to run a bitcoin node since even the current 300GB and the 1MB size is way too huge for them, but when we make suggestions of increasing block size to 16MB or lift the cap altogether, we have to evaluate the technology capabilities worldwide and not just within 5-10 developed countries, just my 2 bytes on the subject.

My mechanic likes to say "you can pay me now, or you can pay me later, either way to fix the problem I am getting paid."

If we increase the block size then yes, running a node will be more money as you will need more drive space and more ram. (pay me now)

If we just use sidechains, (lightning, liquid, etc.) then the data size is still there BUT it's just in another database.
And you have the added complexity of running another app / database / whatever. (pay me later)

And if you and I are not connected that way (I am using lightning you are using liquid) then we are back to settling on the main chain.

It's not that any particular way is inherently better or worse, it's just that there is not a one size fits all solution so they all have their give and take.

-Dave

[mikeywith]

With that being said, not all "poor" people will want to run a bitcoin node since even the current 300GB and the 1MB size is way too huge for them, but when we make suggestions of increasing block size to 16MB or lift the cap altogether, we have to evaluate the technology capabilities worldwide and not just within 5-10 developed countries, just my 2 bytes on the subject.

My mechanic likes to say "you can pay me now, or you can pay me later, either way to fix the problem I am getting paid."

If we increase the block size then yes, running a node will be more money as you will need more drive space and more ram. (pay me now)

If we just use sidechains, (lightning, liquid, etc.) then the data size is still there BUT it's just in another database.
And you have the added complexity of running another app / database / whatever. (pay me later)

And if you and I are not connected that way (I am using lightning you are using liquid) then we are back to settling on the main chain.

It's not that any particular way is inherently better or worse, it's just that there is not a one size fits all solution so they all have their give and take.

-Dave

I don't think I can agree with this analogy, you see, your mechanic makes sense, if your car has a problem - it must be fixed, you can't ignore it, and delying it can only cause more harm, even if you were to sell the car without fixing it, it will be worth its price minus the fix.

On the other hand, the block size isn't really a problem, and even if it was a problem then delaying the fix works, in other words, the longer we stick to 1MB the less the size can get, and then eventually, the equipment needed to run a full node will become cheaper as we move forward, and the internet technology is going to improve (regardless of how fast that may happen).

Now back to why do I think the block size isn't a problem, people are still using bitcoin, they have not abandoned it, bitcoin still works just fine and has more transactions than most other coins combined.

What's even more important than all of this, is what do you expect from increasing the block size? most people I discussed this matter with think that when bitcoin becomes cheaper to transact, more people will use it, and adoption will go to the moon, but I don't see that happening for a few reasons.

1- Other coins with instant confirmation and almost zero fees are not being accepted everywhere and don't have a fraction of bitcoin adoption.
2- The main layer isn't suitable for micropayments and every day's transactions,  you are always subject to wait for hours until your transaction is confirmed, and if confirmation isn't important then people can still do with 1MB blocksize and pay the lowest fees possible.

There are many other reasons why blockchain as-is isn't fit to do the kind of things some people expect, increasing the block size could "solve" some "issues" but it can create worse issues.

[PrimeNumber7]

It would take the equivalent of 58 days for each of those 35 million people to make a transaction with the mainlayer. If only ONE THIRD of the worldwide humans ever wanted to make a transaction with the mainlayer, this would take 4290 days! As an example, a person who pays into his pension every month could only do so every 4290 days! (earliest!)
 
So far to the problem, this is known to us, and why "we" research things like Lightning.

Unfortunately, we now realize that based on the above figures, not even so-called settlement transactions or individual transactions on my Mainnet wallet would be possible, unless I pay fees that are so high that they are no longer in any relation. EVERYTHING that will ever happen on BTC would only happen on the semi-secure second layer which is still in its infancy after 5 years. (Including all of my savings, which I may have set aside for my retirement, would exist for my lifetime only as LBTC, and possibly be managed by a custodial service).

LN channels are opened and closed frequently today because many people are testing and experimenting with LN, and because the number of financial transactions that can be conducted via LN is limited. As LN becomes more established, it will become less common for LN channels to be closed, and there is no technical reason for a LN channel to need to be closed.

Bitcoin adoption is not going to go to 100% of the world's population tomorrow. It likely will be decades before any substantial portion of the world's population is using bitcoin to conduct financial transactions on a regular basis. There are limitations to Moore's law, and it is slowing, however, in the future, the cost of RAM, CPU processing power, and network capacity will decline. My prediction is in the future the maximum block size for bitcoin will be increased in order to allow for more LN channels to be opened/closed.

With LN, you do not need an on-chain transaction every time you receive money or move money to some kind of savings. If you were paying into a pension every month, you would not need to open an additional channel every month if your channels have sufficient inbound capacity (which they should).

LN is also not custodial. If you have an LN channel, you control the private keys that can spend the coin in your open channels.

1. LN is only meant for micro-transaction

LN can facilitate and is useful for micro-transactions, but this is not the only use case for LN. LN can be used for transactions of any size, provided there is sufficient LN network capacity.

[Wind_FURY]

You guys got me wrong. I´m NOT talking about simply increasing the block size, this would not be work in the long therm. If its not the bandwith who limit this, the disc space will do so definitelly earlyer or later. I also didnt want a discussion between big and small blockers here. Its a fundamental problem a lot of people misunderstand and need to be solved in a rational way, ahead of everyones personal financial interests here!

Lightning isn't meant to solve Bitcoin's scaling issues and neither are sidechains. They are only protocols that extend Bitcoin's functionality while maintaining the base layer secure and decentralized.

you absolutelly missed my point. You can´t even use the base layer as you can read in my first post! THIS IS SOMETHING NEED TO BE SOLVED! simply to say this has to be limited, is in the worst case a death for bitcoin at a mass adoption!

and again: to solve the problem isn´t simply increasing a blocksize or anything, its a complex dilemma that needs a lot of research!

think logically!
 

OP, you are not talking about increasing the block size, and you are saying that off-chain payment channels are useless BECAUSE "you can't even use the base layer". Then what is YOUR solution?

[figmentofmyass]

And if you think that is bad, I have seen worse elsewhere.

It's bad, but how low the barrier to entry should be?

it shouldn't be based on developed world standards. i think that much is clear.

Increasing the block has nothing to do with resistance to censorship. This is another misconception that occurs in the Bitcoin community.

In order to introduce censorship in the Bitcoin system, it is necessary to introduce censorship for all mining pools. But the resistance of mining pools to censorship does not depend on the block size.

at large enough block sizes, very few people will be able to run full nodes, and orphaning rates will cause a high level of mining concentration---both factors that weaken bitcoin's censorship resistance.

so in an indirect way, it does depend on block size.

bitcoin's security model depends on full node security, not mere mining security. it is full nodes who keep miners honest. without a vigorous network of full nodes, bitcoin's security would essentially be downgraded to SPV---the blockchain's history would just be whatever miners decide it is.

[DooMAD]

BTC's creator talked about increasing the block size limit as the demand for space increases but he never talked about segwit or a L2 solution.

If you're willing to take Mike Hearn's word for it, satoshi did have some ideas around payment channels, which are effectively the progenitor for the Lightning Network:

https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2013-April/002417.html