Stuart_Armstrong comments on How to treat problems of unknown difficulty