Davidmanheim comments on Using expected utility for Good(hart)