If Romney is right and the polling methodology is poor, won’t the exit polling lead to equally bad results? Or is a different metric used?
If not, then there might be an election 2000 repeat where states are called, then retracted, awarded to the other guy, then retracted.
I think the Romney complaints against polling (other then the fact that it gives a result that they don’t like) is that the likely voter models are incorrect somehow. Given the likely voter model of exit polls is “is the person coming out of a polling station?” I can’t imagine that would be an issue.
If North Carolina goes for Obama, that’s it. Romney has no hope. At all.
Ohio being called for either candidate would probably seal it. But that’s not just a matter of winning Ohio-- It’d mean winning Ohio, and it not being too close to call. Margins matter, for predictions.