I think if this is possible and the incentives can be made strong enough you have a winning coin pool even if there's a theoretical thundering herd worst case. Bitcoin suffers from the theoretical risk of no one mining it yet that seems to be going pretty well.

k00b

Looks like [Johan Halseth had the same idea (at the same time?)](https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2023-October/021978.html) so maybe it's not super dumb.

bitcoin

Scaling Lightning With Simple Covenants

pietro1861

> Anyway, I think that's an interesting way of capturing a big concern
with this sort of approach (namely, "what happens if the nice, scalable
path doesn't work, and we have to dump *LOTS* of stuff onchain") in a
measurable way.

This seems to be a consistent problem with L2s that take activity off-chain via some kind of cooperative utxo yet design uncooperative exits such that they're only unilateral. They reduce N participants to 1 utxo but 1 utxo can turn to N utxos at any time.

Has anyone proposed a design with cooperative subsets exiting in a batch?

I don't know anything about anything, but it'd be cool if these pools were designed such that participants could fork out of them together.