I've said it before, but I'll repeat it now.
People here are too influenced by 2004. The nomination isn't always decided by the first states. The fact that every winner except Clinton has won one of the early states is misleading, because in many of these cases the race went on for a long time afterwards.
For instance, Carter won both Iowa and New Hampshire in 1976 but still had to fight for the nomination all the way to the end. Ford in the same year also won both but had an extremely close battle with Reagan into the actual convention. When you have a situation where one candidate is extremely strong in all the later states and they're coming in one big shebang closely after the other states things may be different. While I'm the first to agree that Guliani has a problem in being so weak in the early states I think calling him done because of it is overdoing it.
I think Gustaf is fundamentally correct. For every nomination decided by early primaries there was another where they didn't matter. The mix of candidates and the primary calendar have to be looked at as a whole. This is a very different field from 2004, 2000, or 1992. There might be some resemblance to the Dem field in 1988 (look up jokes about the seven dwarfs, or given the way some fell by the wayside, the seven deadly sins.) Even if there is a resemblance there is a very different calendar in 2008 and that will impact the field as well.
There are voters in the big states on 2/5 who will watch the earlier states, but I think it will focus more on gaffes and collapses, not on specific placing in the early states.