There are two main reasons that presidential nominations are much harder to understand and to predict than other types of elections. One is that there are relatively few iterations of the process, and since that process keeps changing, it's never easy to know which past examples are relevant or how they are relevant. The other is that most of the time, most of the key independent variables tend to vary together: that is, the candidate doing well in the polls is also the candidate picking up endorsements, raising money, having a good organization, and otherwise doing well on all the things that might possibly be important. There's been some variation, thus allowing people to study it to some extent, but it's tough.
But this year! Nate Silver remarked somewhere recently about the split between endorsements and poll ranking, and now Mark Blumenthal, using his great Power Outsiders survey, shows that there's yet another split on organization. Polling and endorsement also-rans Michele Bachman and Rick Santorum may wind up with the best organizations in Iowa; at least, that's the perception from Blumenthal's respondents. And then there are splits in fundraising, and in the candidates themselves, or at least their experience/qualifications going in. The bottom line here is that most of the statistical tools available work a whole lot better (and should work a whole lot better) when the possible explanations for something vary in lots of crazy ways, and for whatever reason it appears that we're really getting a ton of that this time around.
Big, huge, major caveat to anyone thinking of doing this research: there's a huge difference between explaining: who will win Iowa given differences in endorsements, polling, fundraising and organization up to the day of the caucuses; who will win the nomination given those differences; and who will win the nomination overall. All are worthwhile questions, but they're different questions. Note in particular that one serious candidate from this cycle, Tim Pawlenty, was defeated back in August; others (Barbour at least, and perhaps Palin, Thune, Daniels, Christie, and more) tried to win it to some extent but either gave up or were defeated before getting to the full candidacy phase. Any analysis of winning the nomination that ignores Pawlenty and the other on-paper seemingly plausible candidates is, in my view, going to be missing a large part of the story.
But putting that aside, I think all of those who research nominations owe an enormous debt to the 2012 Republican field.