Maybe an IESG statement on this respect can help here.
Is the existing text in RFC 5226 not sufficient? It contains extensive
text about the purpose and role of designated experts, and was revised
substantially the last time around to try and find a good middle
ground between being overly prescriptive and giving experts a "blank
check" to do what they want.
Nothing in the discussion I've seen so far in this thread seems at
odds with or beyond what is already in RFC 5226 (but I may be biased).
Thomas