[Accepted] Amendment to SE-0466

I think this is related to a question I asked a while back of why we allow a protocol definition itself to have actor (or sendable) requirements, rather then only the individual protocol requirements: