[BUG] Investigate collator rewards when session has unpredictable block count

[1xstj]

We have observed that sometimes DKG can lag with key upload, also considering that session does not rotate until dkgkey and nextDkgKey are updated this can lead to different block counts for some sessions.

Our collator reward logic depends/assumes all sessions having an equal number of blocks, we need to investigate what sideeffect this new session rotation logic has on collator rewards and maybe redesign the rewards logic.

[1xstj]

Adding more info, as part of parachain staking pallet we now store RoundInfo, https://github.com/webb-tools/tangle/blob/48a8935a0d5ec26e72c1a90b04b03697e5e9179f/pallets/parachain-staking/src/types.rs#L1576

This has details on which block the round started and when it ended, perhaps this could be useful here.

[drewstone]

The deliverable for this @1xstj should be a document that clearly explains how the rewards work and ends up in the documentation sites.

[1xstj]

Closed by tangle-network/webb-docs#85