It's a little different with proof of stake. If there's a minority group that holds a large stake that "attacks" the blockchain against the majority's wishes, the majority group can decide to roll back and blacklist the minority group. It's much more difficult to blacklist hash power in this same way.
>the majority group can decide to roll back and blacklist the minority group. It's much more difficult to blacklist hash power in this same way.
and that is supposedly advantage of pos over pow?
In philosophical view the pow -> pos means taking power away from the more technical and giving it to the more business oriented group/approach. That has happened to Internet for example, and I just personally dont like such transitions.
There is a difference (with the introduction of slashing and user activated soft forks) in 51% attacks that revert finalized blocks. In PoW this type of attack would become cheaper as it persists, in PoS we have some defense that keeps making this attack expensive.
On top of that the costs of attack are much greater in PoS than PoW, making these attacks much harder to begin with.
Vitalik and company have zero history in the proof of stake game, and what little security history they do have is pretty much awful. So I'm not too worried about what they have to say about things. If it's not something he copied from Tezos or ADA, I wouldn't consider it relevant to the discussion at all.
