I think Ordinals and BRC20 may stay.
Of course they will. It was unclear during the first months, but now, enough time passed, so that everyone can see: nothing will be changed. Some people may want to censor them in that way or another, but eventually, all of their transactions will be confirmed, as long as enough fees are paid. Which means, that the mainchain will be turned into a spamchain, where everyone will include all non-payment things, and treat it as a P2P cloud storage. This is where the mainchain is going, and that won't be changed, because people didn't stop it during early stages, so nobody will do that now.
They argued that they generated money for bitcoin miners which are maintaining the security of bitcoin network.
Of course. And they also started bloating the UTXO set, which means, that there is more and more disk space needed, to run a pruned node. In the extreme case, the size of the chain can be comparable to the size of the UTXO set. Fortunately, we are far from that, but now, everyone can see, how to spam the chain. So, that kind of things can be done: blocks can be filled with a lot of outputs, sending zero satoshis to the UTXO set, or a lot of non-consensus data in the witness space, behind OP_RETURN, or anywhere else.
And of course, when there are less payments, than needed to fill a block, it costs nothing for miners to spam the chain, so some mining pools may start doing that in the future. Then, blocks will never be empty again, because unused space can be always filled with non-payments, for no additional cost (and then, non-mining nodes will pay for all of that).
What do you about this?
I think if someone wants to focus on payments, then processing everything, what is happening on the mainchain, will no longer be profitable. And then, spammers should be left, where they are. Those, who are focused on payments, can start building subnetworks, where rules will be more strict, and where only payments will be allowed. Everyone is free to use the original spammed chain. If it will be too costly, to send anything from the official node, then people will switch to these subnetworks. As long as the current level of spam is not kicking nodes out of the network, nobody will care.
I do not think Ordinals can be eradicated even if some bitcoin developers will try to do it.
I think they can be blocked from some subnetworks, while being allowed on the main network. Nobody can stop miners from confirming spammy transactions on-chain. But nobody can also force nodes to use the official software, and process spammy transactions in the first place. Which means, that there will be a place for both, at the same time, and users will join the network they want. If spammers will want to turn mainchain BTC into BSV, then they will. If miners will want to activate 1 TB witness per block, through some soft-fork, then they will. Segwit told everyone, how to increase the size of the block, if some users will decide to make it unlimited, then it will be done. And then, it will be up to the users, if they will still want to use the official software, or not.
I guess next limits will be lifted, one by one. Then, developers will be responsible for nothing, and everyone will be able to deploy absolutely everything. And then, there will be a lot of spam, while some people, caring about payments, will form their own subnetworks, when they will focus only on transacting, and ignore the rest of the traffic (because you cannot force old nodes, to validate all new rules, which spammers would want to introduce in the future, even if they will have hashrate majority on their side).
Edit:
If an inscriptions-focused group or a censorship-focused group forked Bitcoin Core to change the rules, they've made a new chain.
There is no need to make any "fork", if you want to process only a subset of the mainnet traffic. The same chain can be used, and the heaviest chain of the Proof of Work headers can be always followed.
For "inscriptions-focused group", they can increase the size of the block, just like Segwit did, by making yet another soft-fork, where blocks would optionally commit to data, in additional space, processed only by new nodes.
For "censorship-focused group", they can process the subset of the mainchain traffic, and ignore the rest. They don't have to process or store any transactions, that they don't want to. They only need a minimal setup, to make sure, that blocks are valid, and if one node can make a proof, that some data between offset X and Y is valid, then the rest of the nodes can accept the proof, validate it, and move on.
Practically: no--not without a major consensus change that the ecosystem would have to agree on.
Nobody knows, what is the exact code, which is running behind some node. People can only check the end results. And as long, as other nodes are working on the same block headers, they can store nothing behind them, if they want to. One group of nodes can mine soft-forked blocks with 1 TB per block. Another group of nodes can mine empty blocks, with only coinbase transactions. They can live on the same chain of block headers, as long as they all agree, that a given chain of Proof of Work headers is the strongest one.
Core devs can tweak relay policy for network health, but they can't, by themselves, "censor" valid transactions.
They can write the code, which will optionally process the subset of mainchain transactions. But as long as the whole spam can be handled by all node operators, nobody cares, and nobody would care.
