I think it was more of a regret that his intuition about the problem area gets invalidated so quickly. The top technical guys are supposed to have good intuitions about what works and what doesn't; this is a major part of the value they can add. So I take this as a very serious self-criticism.