This might be a case where there's overlap between ad-hominem attacks and a relevant critique of the argument itself. I interpret the comment as: "if the autho

This might be a case where there's overlap between ad-hominem attacks and a relevant critique of the argument itself. I interpret the comment as: "if the author can't follow his own guidance, or if the author's own compliance with this guidance is subjective, the guidance may not be given in good faith". The comment is not just saying asshole=wrong. It's saying that Anil Dash may be a hypocrite, and thus unable to propose good social policies. The discussion (to me) seems more moral or social rather than technical. In cases of moral or social judgements, are you sure you completely reject ad-hominem points? It seems to me that a fully Bayesian perspective requires considering ad-hominem attacks; if the source emitting the message is considered untrustworthy, should we discard that information?