Yeah, I didn't want to be contrary just for the sake of it, the heuristics you mention seem like good ones, and if followed would probably already cut down on q

Yeah, I didn't want to be contrary just for the sake of it, the heuristics you mention seem like good ones, and if followed would probably already cut down on quite a few superfluous references in most papers.