It's worth trying to improve ETH client diversification so no single client is above 66% but let's not forget Ethereum is the only L1 that even have multiple clients.

Sometimes Ethereum is judged by weird standards relative to everything else because it's doing hard things.

Isn't it more than "worth trying" for to get it under 66 percent? The penalty is flat out severe with a supermajority. I hope this issue doesn't start to get downplayed. I saw CB's announcement this week. But it also went back up from where it was a week ago or so.

am thinking about this with Farcaster a lot. There are multiple clients but all of them are based on the same node client (Hubble).

though Farcaster maybe too early to have that level of Ethereum client diversity?

Isn't it more the fact that we went through a lot of effort to have a multi client system, so if there is a majority client and it has a bug then we are no different to a single client chain? If you're the best you don't compare yourself to the second best. you compare yourself to yesterday!