💎 Issue
Background
We have been investigating the issue with aggregates dissemination and in particular this weir bimodal aggregates arrival distribution.
It looks like Prysm plays some role in the introducing of the second peak
Description
Prysm filters aggregate messages well for the current slot, i.e. it IGNOREs the messages containing seen aggregation bits. But as soon as the next block is imported the filter looks to be gone because all the older aggregate messages are accepted.
I've made the branch with debug println and here is the slice of its output