Thank you all very much for your time and constructive information.
Some of you, said, and are I guess right, to say that I'm not sharing enough details in my question.

Thing is that, I'm usually quite good in thinking and delivering but I wouldn't want to self-advertise too much and sound like I brag, and ultimately developers are hard to assess and judged mostly by result. I work in a startup, have written C++, Node.JS, GPU code (mods to OpenCL mining code too), and I have ideas, but I either see them implemented, or they are too hard to just start implementing.
 
I wanted to get some pointers, and I've got quite a few.

An ICO is something I'd like, by doing something useful too, but I'd rather go somewhere where it's not already stepped, or join a group.

One idea of mine:

I want to create a decentralized coin that consumes so little disk space that a full node can even run in a mobile phone.
This is not just fantasy: I've read quite a few papers and I'm aiming at dropping full history by using hashes (edited: not checkpoints, as it should be purely decentralized), and cleverly sharding states, with Merkle queries, and with a possible trade-off in security, (regarding long-range-attacks, but which is not an actual trade-off, if I can mitigate the way I think I can).

But this is all a little overwhelming, my time and energy (as stated) are not entirely optimal at this point.

Generally, I'm thinking how I can contribute with something useful rather than dig for gold, but this is not always easy.

Collaborative projects may be easier to handle, or bounties, or jobs.

But, yes, one potential goal of mine is to be able to earn competitively, so that I can leave the classical industry behind and work for decentralized / bitcoin projects, which I consider more fun!

A skilled engineer would be - to some - somewhat valuable especially if you have a computer-related degree and a mathematical background.

If I knew that it would pay off I would sure launch a coin/token, because it's the sane thing to do, but I don't know and I suspect that it has risk to mitigate.

How can it be done without too much risk?

We could assume that the engineer has good traits, but then again, launching something requires a set of other skills, like marketing, peoples skills etc.

The question is mostly focusing on the how.
How can a skilled engineer make money in bitcointalk?

I'd like to think that trying to launch yet-another-coin is saturated.
Any ideas?

Thanks for answering this. I'd like to know more. Are there specific coins that are well or less well known?

Thanks, this is helpful. I am actually asking for both. Any way to reduce disk space, that can be reliable of course.
I think that decentralization is more meaningful if all nodes are full nodes.
At the same time, if full nodes can safely discard data, in whatever - proven - way, I'm all for it.

Thanks. Are there known coins where this is applied recurringly?
Is there any known impact in security?

Sounds helpful. Thanks! I am wondering about the off-chain solution and whether it could impact decentralization in any way.
Is there any live implementation of sharding somewhere?

+1 to coinbase. Just expect a verification process so don't wait till the last moment if you are in need

One of the problems in cryptocurrencies starting with bitcoin is having to store the (large) blockchain which makes decentralization harder.

Which cryptocurrencies and technologies tackle this and how efficiently?

How effectively is sharding and/or pruning applied?
Any experience?

Has the idea of voting the next miner been proposed in an altcoin?

E.g. Candidates participate in a recurring contest. Voters, who can be e.g. currency holders, vote for a winner.
The winner adds the next block to the blockchain and receives the reward.

Proof of winning the contest could possibly consist of a combination of legit votes as signed messages