We already made steps towards allowing users of PARSEC to use an alternative to supermajority as a premise for describing an interesting_event.
Make sure this logic is airtight as it needs to be correct for integration in routing.
For the proofs to hold, we probably want to ensure that in the case where a single observation is sufficient to initiate consensus, any event that sees that observation behaves as if they also were carrying the observation.
Also, we need to make sure that no decision that pertains to network toppology may use this customizable definition of interesting event.
This task is critical priority as it stands on the hotpath towards routing integration.