Yes. It's like the government of the swiss says, "dear people of swiss, our constitution actually demands a referendum for us to do this properly. But since we don't trust your decision, we decided the best-solution is to tricky-track around the problem, so that we can do it without a referendum. That's brilliant. It ensures the system's stability."
The same is with RBF, no matter if opt-in or full. If it goes through, it forces every node to confirm the existance of rbf.
I'm really curious how many people will upgrade the next core version. It could happen we'll find bitcoin soon in some state of ungovernability, like belgium, where even the leading party (=core) is unable to push needed things forward.
While I'm for the decentralization of development, I'm not sure if such a state is better than what we have now.
I am adamant that it is a superior state to what we have now, if we cannot achieve such a state I would lose much of my interest in Bitcoin and I would consider the governance mechanism of Bitcoin to have failed, this is the first real test of this mechanism so the outcome is hard to know, since there is no historical precedent to this situation. This is a experiment in decentralized governance.
I worry such a state could destroy or paralyze bitcoin. But I agree with you. If voting with the client is a sin, bitcoin's protection against malgovernance will have failed.
After all if we had to be reliant and dependent upon a centralized authority in the form of Core in order for Bitcoin to survive then this would defeat the very point of "trust" without centralized authority. I think that these alternative implementations will increase the blocksize limit more and faster then Core would.
However I do suspect that under a more decentralized model of governance that it could slow down development in other regards. Since implementations would need to negotiate any consensus critical changes with other implementations and the wider community more then they presently would. This could have the effect of slowing down development though it would also make development in Bitcoin more conservative, which I would actually consider a positive development.
Maybe longterm bitcoin will be made more modular, so that only very few rules are consensus critical. In such a state it would be impossible for any party to change the 21M limit, which would be good for Bitcoin as a store of value.