Consider RLN and rate limit manager within Messaging API

[Ivansete-status]

I recently thought: why is RLN compliance part of Reliable Channel API, and not Messaging API?

If a network has RLN enabled, then one can't publish messages on it without RLN membership, and can't verify proofs without access to smart contract. In such case, Kernel API (relay/lighpush/filter/disc) still make sense, but Messaging API becomes useless.

So I wonder if it would make sense to implement it in Messaging API instead. Or am I not seeing something here?

Originally posted by @igor-sirotin in logos-messaging/specs#104 (comment)