Jason asks a good question over at The Plank: Will a Romney win in Iowa matter if Huckabee takes second? The theory is that there's only room for so many stories out of Iowa, and Huckabee's surprise second could overshadow Romney's victory.
This is, not surprisingly, the scenario the Giuliani campaign is banking on. And if you combine Huckabee with whatever happens on the Democratic side, it starts to look pretty plausible. That said, I think Romney will benefit from anything other than an extremely narrow victory in Iowa (under five points) - or, obviously, a loss. At least if the race looks about like it does today in New Hampshire.
Here's why: According to Pollster.com, Romney is currently up by about six points on Giuliani in New Hampshire and 14 over McCain. (Huckabee's about 22 points back, though it's important to note that the Pollster.com chart is a bit of a lagging indicator. He could be a little closer by now.) As long as neither Giuliani nor McCain makes much a showing in Iowa (as appears likely), I think a Romney victory there--even one that doesn't generate a huge media splash--basically reinforces the status quo in New Hampshire, which means Romney wins fairly comfortably there, too. And if Romney wins both Iowa and New Hampshire by reasonable margins, things start to look very good for him.
Or, put differently, Giuliani needs either a close second in New Hampshire or an outright victory there to slow Romney's momentum. If Romney finishes first and Giuliani a distant third or fourth in Iowa, I don't see Giuliani turning around and gaining much ground on Romney in New Hampshire, rergardless of what Huckabee does to Romney's media coverage heading out of Iowa.