I think outlining the fallacy of applying moralistic thinking ("well people just SHOULDN'T DO THAT") to systems design was the more interesting part of his post, though.