So BitVM has that issue? If I'm understanding correctly, Bitcoin L1 can't verify ZK proofs at consensus level, so anchoring relies on operator good faith?
If that's right, then maybe the more natural approach for Bitcoin L2s is to compose only from primitives that L1 can directly verify. Ark protocol seems to fit that mold - periodic batch settlements using only what L1 natively understands, with each settlement providing on-chain finality at fees split across all participants.
Curious whether you've looked at how Ark handles this differently, or if you see issues with that approach too?