I'm not a pollster, but I think this PPP commentary misses something big.

Around the time of SC, GA, MS, turnout among black voters and young college students wildly exceeded expectations.  That's the primary reason that Obama outperformed his polling expectations, I would think.

Now I don't have any inside baseball knowledge, but I would have to think that the pollsters would have upwardly adjusted their expectations on turnout among black and young voters, following their big misses (which PPP mentions).  But as enthusiasm for Obama waned a bit, and as college students went into finals (and now summer), that unexpected turnout might be expected to drop.

I think this is what happened in the recent states, particularly in Ohio and PA.  Pollsters adjusted their models to reflect higher turnout among Obama groups, but these groups started seeing more normal levels of turnout, explaining why the pollsters got things so wrong in recent states.

I suspect we'll see the same thing in NC and IN.  I'm fully expecting to see about a 5 point "bump" in each state from most polls in favor of Hillary.  Maybe after Tuesday, pollsters will adjust their models back to reflect the dampened ardor for Obama.  But in the meantime, I think I'd expect to see a sharply lower college turnout, and a slight increase in rural working class voters.