It's down to the wire (PDF):
I'm sure Paulites won't like to hear this, but whether or not Ron Paul wins Iowa, he's not going to be the Republican nominee, so I'm mostly interested in two things: first, where Mitt Romney finishes (both relative to other candidates and to his 2008 performance, when he got 30,021 votes for 25% and a second-place finish); and second, whether any of the Not Romneys breaks out from the pack. (In the event that there is a brokered convention, Ron Paul's performance will suddenly become relevant as his delegates would could be a significant bloc, but otherwise, he's irrelevant.)
Based on these numbers, Mitt Romney seems to have the opportunity to grow his support somewhat, perhaps as high as 30% or so. Michele Bachmann seems to be headed towards Jon Huntsman-style irrelevance, and given Newt Gingrich's rapid descent, Rick Santorum and Rick Perry seem well-positioned to emerge as the leading Not Romneys.
Unless Gingrich somehow turns things around, the battle between Perry and Santorum will be the most interesting contest of the final few days; if one of them breaks away from the other, he could pose a significant challenge to Mitt Romney. If they remain deadlocked, however, it clears the path for Romney to waltz through a divided field. On paper, Perry seems to be better positioned to pose that challenge, not just because unlike Santorum he's won his most recent election and has demonstrated superior fundraising ability, but also because he is likely to fare better than Santorum in South Carolina, which comes up immediately after New Hampshire. If he's able to make that argument, he could stage a surprise comeback; if not, Santorum's momentum could carry him to be the next top Not Romney.
Sunday's Iowa Poll from Selzer & Co. and the Des Moines Register will provide another big clue, but a lot can change between now and Tuesday.