The power of miners to censor the blockchain just means they can control how it behaves within the rules of the existing protocol. Changing the protocol itself is something miners have no ability to do, regardless of how much hashing power agrees to it (since miners can’t force users to switch clients). Traditionally protocol changes have only been done through consensus, and I am very concerned with how the community is rationalizing the fact that the current hard fork is being forced through via majority rule, even going as far as calling it a “feature not a bug” and considering hard forks as a means of legal intervention a sustainable way for the community to self govern and prevent errors in the future (and then the worst one: “it’ll just happen once”). It is also seems quite clear that at least some hard fork proponents support the implied attack against the old blockchain in order to secure success of the new fork — the precedent of organizing attacks to enforce contentious hard forks is probably the one I’m most afraid of.
At this point I no longer oppose the hard fork since the vocal majority has already made up their mind of abandoning consensus and simply forcing their desired changes through. IMO the act of abandoning consensus was the biggest danger in the first place, so since the damage has already been done we might as well go along with it.
But I don’t think there should be any illusions that this has been a major disaster for ethereum that can’t possibly be spun in a positive light. First the dao hack ruining the reputation of turing complete smart contracts, and now the contentious hard fork ruining the reputation of ethereum and turning the community on itself. Of course in the end we’ll probably be fine, but I think at this point individual projects and dapps should be a lot more careful about relying on the ethereum protocol… Hiveminds can be very unpredictable.