MSNBC's First Read makes an interesting point about the way to interpret caucus results (scroll down to the fourth item):
Clinton’s super problem: By our count, the Clinton campaign hasn’t publicly announced the support of a new superdelegate since just after February 5. Indeed, since Super Tuesday, Obama has gained 47 new superdelegates, while Clinton has lost seven (including Eliot Spitzer). Does Clinton have a bigger problem on the superdelegate front than folks realize? Why do we think party leaders -- who saw the Democrats lose governorships, state legislatures, and the control of Congress during the Clinton years -- suddenly jump on board the Clinton campaign? Isn't this the reason the Clinton campaign has only been able to keep uncommitted supers from climbing board Obama's bandwagon but they haven't been able to woo a new super to their side in a month? ? Isn't this also an explanation for why the Clinton campaign has done so poorly in the caucuses? The caucuses are made up of the activists who follow this stuff closer and think about things like electability and who can help the party keep Congress, etc. If Clinton's not winning over caucus activists, why should we believe she'll win over a large enough chunk of superdelegates to overcome Obama's pledged delegate lead? Ultimately, her best chance is to convince supers that Obama is completely unelectable on par with McGovern, an argument that might have been helped a tad by Rev. Wright. [emphasis added.]
It does seem a little tricky to argue, on the one hand, that you do badly at caucuses because they're undemocratic but that, on the other, you're going to clean up in the most undemocratic "caucus" of all (i.e., the superdelegate caucus). Obviously, the disproportionate influence of party activists isn't the only reason Hillary has struggled in caucuses. But I'd guess there's something to this.