If the GOP is tied with anything involving the MSNBC folks...
When you put in real actual voter turnout using a model of 2010 or 2004, Romney is going to make everyone at MSNBC start crying.
I think that a D/R/I model based on 2010 would be just nuts. Instead, something about halfway between that and 2008 is likely reasonable.
Though the “enthusiasm” of the D voters is definitely lower than 4yr ago, their turnout machine is clearly getting a fair number of -bama voters to early voting, so even though they are “more sporadic” and “less reliable” (CBS Radio news terms on story just now) that machine is cranking and producing.
Meanwhile, the Tea Party has not gone away, either in raison d’etre, numbers, or enthusiasm. They will turn up, both to vote and to crank the GOP machine. Since TP was non-existent (or even negative) in ‘08, the gap will be greatly narrowed.
Halfway seems to be a reasonable place to be, so close to a completely balanced 33/33/33 mix seem about to me at this time, adjusted for each state’s tendency. Since Romney is easily winning the indy’s, that yields an easy win for him in most states.
Oh- one more thing: I suspect that there are many more people falsely stating “I’m voting for Obama” than those who are falsely telling pollsters “I’m voting for Romney”. I believe that has been reflected over and over again in crosstabs via “I’m conservative and voting for -bama” and “I’m Republican and voting for -bama”. That would give a systematic (though small) false blip up for him.