I suspect these are conflicting goals but an interesting question. My idle pondering are on ISAs that minimize the challenges in wide superscalar OoOE.