It seems to me that, as blockchain scaling solutions are doing, to deconstruct an Ethereum blockchain (the Ethereum format is the standard now) into its constituent components (data, programme, execution, state, state transition, etc.) and to try to figure secure ways to scale by having those components in different places while talking to each other through proofs and subjective/objective tradeoffs is an intelligent way of approaching the problem.
I think all the new systems that are evolving, including ETH 2.0, Cardano, Polkadot, Dfinity, channels, Plasma, rollups, etc., make deliberate choices.
The key is to understand those choices and to not be confused into believing that they are not choices but “solutions” to scaling, security, bloating, backward compatibility, etc.
In my opinion, Ethereum Classic (ETC) is a blockchain that makes the choice of remaining fully integrated and deliberately unscalable as a tradeoff for being as secure as possible.
In this context, “secure” means not disaggregating the Ethereum standard components, but remaining fully integrated and fully replicated to minimize the risks of the components talking to each other through less secure or not battle tested contraptions.
I do strongly recommend ETC to integrate all the tech necessary to interact with all those L2,3,4,5 systems.
That is the other key for ETC’s success.
A roadmap that is consistent with the vision I propose above would be:
1. SHA 3
2. FlyClient/NiPoPoWs (I don’t know their popularity, tradeoffs or compatibility, but the idea is interconnectivity with other systems)
3. zkSANRKs (not only for privacy, but also for proofs as above)
4. EVM and Account Versioning for backward compatibility
5. Permanently fixing the block size at 4 million or 8 million GAS (so miners cannot enlarge it further as happened in the Ethereum network)
6. A limited in time treasury to finance the above and re-float ETC (knowing it is centralizing and needs to go away)
7. Eventually eliminating checkpointing and MESS
Code Is Law