Another line of thinking I've had about this genius Nate Silver is:
-He predicted 49-50 in a single election. Big whip, did anyone get less than say 45-46 states right? It was a landslide election & it was no surprise that it was on election day. At best, Silver simply weighted the most pro-Democrat polls that year more heavily to get to the narrow predictions of Obama winning Indiana and North Carolina. Other than those 2, were any others really in doubt in 2008?
I'll take his predictions over that of Rasmussen & Gallup when he has managed to be right in elections consistently over 2 decades. One election prediction doesn't make it some full proof proven model by any means. I have a feeling Mr. Silver will be finding that out in a week