I spent last evening at Maine caucus training (run by the MDP, not any particular campaign) and since then, I've been thinking that all the hype about big turnout, particularly in the 18-34 year old, may not make much difference. Why? The pre-determination of delegates.
See, if Iowa is anything like Maine, the number of delegates each precinct has to divide has already been set, determined by the level of partipation in the 2000 caucuses. If turnout increased equally all around the state, then, yes, that may be of benefit to Gephardt and Dean, with their GOTV operations.
However... If turnout is increased only in urban areas or college towns, but remains rather average in rural communities, then those urban areas are merely dividing up those predetermined pieces of pie, not picking up anything from other counties or even precincts.
It is in the rural areas where the "momentum" of Kerry and Edwards may trump the Dean/Gephardt organizational machines. Every precinct gets at least one delegate, and even if only one person shows up, that delegate preference has as much weight as the 500 people who might determine a single delegate in Des Moines. It's not dissimilar to the inherent inequity in the electoral college process - even states with tiny populations get at least 3 electoral votes.
I still think the key, however, are those predetermined delegate counts. Unless the groundswell for Dean or the union turnout for Gephardt is spread equally over all precints, and I doubt it is, all those new caucus goers in Iowa City and Des Moines may find their power mostly diluted.