Fantastic book (I just read it last week); fantastic discussion. I submit that the "polymath" theory is flawed for reasons not fully explored. First, gleaning "truth" by comparing competing versions of events is time-consuming and complex. Truth was a lot easier to come by when we could be reasonably sure that Uncle Walter was speaking it. More importantly, however, the whole premise that truth can be gleaned in this manner is false. That is also the premise behind our "adversarial" (as opposed to inquisitorial) legal system, and after 17 years of practicing law, I can tell you it does not work. I believe the reason it fails is described beautifully in the book - as issues become too complex for non-experts to reason out, we become reliant on experts to interpret. Then we end up with competing experts, and no personal basis (i.e., education and experience) to tell us which one(s) to believe. So, we choose based on "peripheral" clues - who tells a better story, has whiter teeth, whatever - that often lead us astray. Yeah, for something pretty straight-forward like urban legends of headlight-flashing gang killers, Snopes works. But try "fact-checking" assertions about the economic impact of the farm bill, or why gas prices are high - you can't, because it's too complex. We need authoritative impartial sources who know enough about the subject matter to ask the right questions, report the right answers, and take positions when the evidence demands it. When Cronkite took a position on Vietnam, it was as good as over. No one has that kind of credibility today, and "melding" a Bill O'Reilly with a Michael Moore is no substitute.