In late 2003, I made myself look smart by saying publicly that I didn't think Howard Dean would win the Iowa Caucus, despite his lead in the polls (roughly 30% in Iowa, with others well behind). Dean was very different from the other candidates, I argued, and people had pretty strong opinions by that point on whether they wanted that kind of candidate or not. Anyone who wanted a populist, Washington-outsider, opponent of the Iraq War, was with Dean. Everyone else, it stood to reason, wanted a moderately liberal Washington legislator who had voted for the war -- but hadn't decided which, among the many very similar options, to choose (Kerry, Edwards, Gephardt, Lieberman). Ultimately, the 70% who didn't want Dean would converge on a couple of those other candidates, likely giving one of them the victory.
Today, as hermetically-sealed candidate Mitt Romney emerges to participate in his first debate of the cycle, the Barnstormin' Mormon is polling well ahead of the rest of the GOP field -- nationally, and in early voting states.
The question is, is Romney so different, and well-established, among Republican voters that these polls mean that he's being solidly rejected by roughly 75% of the party?
That's not necessarily what's happening. He could still be unknown enough that people aren't willing to say yes to him now, but likely will later. He could be similar enough to other choices that he'll get their vote when their first choice turns out to be a dud, or not viable.
Polling probably can't answer this question yet. It's still very early, and the rest of the field is largely unknown -- so, for instance, Romney may get a lot of "2nd choice" nods from voters who just don't know who any of the others are. Even favorable/unfavorable numbers don't necessarily tell us whether people will ultimately be open to voting for Romney, once they learn the flaws of the other candidates.
But it could be a bit of a Dean scenario. The Romney brand has been marketed heavily for about seven years now, particularly to the small universe of early-state Republican caucus- and primary-voters, so one might imagine their opinions could be pretty well set. And he is arguably a significantly different product than the others -- more corporate, less Tea Party; less passion, more PowerPoint.
You could make the argument that there are Republican voters who want a respectable businessman with dubious commitment to core conservative principals but the concensus best chance to beat Obama, and they all already show up in polls as Romney voters. That would mean that the vast majority of GOP voters don't want that, but don't know yet what they do want. But it ain't Romney.