Possibly wishful thinking on my part, but I'm on board with Todd Beeton's suggestion of a "unity tour" (which I'm calling a "victory lap") by Barack Obama and Hillary Clinton over the remaining primary states. Why? Because I think the Democrats could learn an awful lot about the salience of Obama's various electoral strategies based on the sort of response they got.
The next state up on the primary docket is West Virginia, which is quite possibly the worst state for Barack Obama in the entire country. Think about the factors that correlate with Obama success: black voters, highly-educated voters, younger voters: none of these things are in abundance in West Virginia.
Say that Hillary Clinton endorses Barack Obama but nevertheless spends several days touring West Virginia with him, and encourages voters to go to the polls and punch their ticket for Obama. Say that turnout is in fact relatively high, and that 70 percent of West Virginians go ahead and vote for Obama -- in a state that Obama would/will lose 70/30 in a competitive campaign. Would that not be a sign that the unity ticket might work? That a lot of voters who might otherwise not vote for Barack Obama might consider doing so with the strong blessing of Hillary Clinton? That the party will in fact unite behind the nominee?
Say instead that turnout is slack, and that Clinton actually picks up a majority, or a substantial minority, of the vote. Not good news for Barack Obama. But it would tell Obama something: there are certain types of Democrats -- rural and working-class whites -- who just ain't going to vote for him. Maybe he just doesn't click with them. Maybe they have a problem with a black guy. Maybe they think he's a snob. This would be a highly useful thing to know. And Obama could instead carve out an electoral strategy focusing on wooing independents and adding states like Virginia and Colorado to his column, acknowledging that McCain is going to get his share of Democratic crossover votes.
The fundamental strategic choice that Obama faces whether to try and win by turning out and unifying the Democratic vote, and taking advantage of the Democrats' substantial edge in party identification, or instead running as more of a 'post-partisan' candidate that works from a different electoral map. The former strategy might well involve selecting Clinton (or a close Clinton ally like Ed Rendell) as his Vice Presidential nominee; the latter might involve someone like Jim Webb.
A "victory lap" with Clinton might be the perfect experiment to test which of these strategies is superior -- and West Virginia would be the perfect laboratory for it.