I think it is great to question Schneier, for the same reason we do security audits. You need to examine your trust anchors from time to time. So for Schneier, we ask ourselves a simple question, is the advice sound? The idea that the math is sound, but the implementations are broken has been a constant refrain from everyone for decades. There would be no benefit to paying or threatening Schneier to make him say that, as it has been his refrain for ages. Tomorrow the people in the field with the skills to attack the math will do so, just as they were doing a year ago. The people with the skills to attack the implementations will do so like they did a year ago, but with a few more obvious targets. We'll all continue to harden our security the same way we have for ages, by assuming that there is a flaw and that it must be found.