The Edwards campaign is trying hard to generate a sense of momentum in Iowa. I recognize a few trusty old dodges, mirrors, and other tricks from the spinmeister's toolbag in their spiel; they are pulling record (even Goldwater-esque) crowds, internal metrics like web visits, sign-ups, bumpersticker hand-outs, and my favorite of all "volunteer calling" all show growing support, etc., etc. (The Atlantic's Marc Ambinder falls hook, line, and sinker for it here.)
I'm more doubtful. It reminds me of Howard Dean's 41,000 committed and identified supporters on the eve of the last Iowa caucus. Still, an Edwards surprise win in Iowa is not impossible. Polls show him in the hunt, and he did well there in the past. He is a good campaigner. Hillary is on the rocks. But it is easy to mistake the final enthusiasm of your supporters, especially if they suspect the end is nigh, for a surge. We'll see what the world looks like right after Christmas.
Meanwhile, my gut tells me the Huckabee surge is fading now in Iowa. It might not be so bad for Romney after all. A narrow loss will look like a comeback after the original Huckabee polling leap knocked Romney's out-of-control expectations back down to earth. Plus, my Iowans tell me the new Romney comparison spots are starting to have an effect.