Our target is a 2-chain interop cluster. The objective is to consider whether consolidation as currently implement remains fault-provable in a reasonable amount of time for the most expensive state transitions under cannon.

Consolidation introduces additional cost factors, compared to pre-interop derivation, including receipts fetching and canonical block lookup. It'll be worthwhile to consider the following, non-exhaustive, cases:

• block dependencies that are mostly filled with executing messages and/or receipts.
• future projection of the earliest canonical block query. And whether the canonical oracle can satisfy such queries in a reasonable amount of time.