We have a test for the super root provider covering the intermediate steps between two super roots. The previous super root is known and the next one is not found. The test asserts that all steps should be the InvalidTransitionHash.

https://github.com/ethereum-optimism/optimism/blob/develop/op-challenger/game/fault/trace/super/provider_test.go#L228-L240

We need to confirm this behaviour with op-program's expectations. I suspect that it should only transition to the invalid hash at the point where the first chain doesn't have sufficient data on L1 to derive the next block - so some intermediate steps might be valid.

This may work out ok for supervisor because it can return a super root for unsafe blocks (although there are potentially failure cases where the sequencer for one chain hasn't produced unsafe blocks even though other chains have safe blocks at that timestamp).

The same test is being ported to the Supernode version (#18617) where the times we get not found may be different so will need to consider that unsafe blocks don't exist.