I agree with you that all proposed L2 solutions based on covenants are horrible (except for Mercury statechains but it's not fully trustless so I guess that makes everybody ignore them?).
I don't understand the last paragraph. What is this research you're talking about? I want to be excited about something too.