From Ambinder, good breakdown on the numbers and what can happen
http://marcambinder.theatlantic.com/archives/2008/02/inside_delegate_math_the_numbers.php
Ohio: Clinton wins by 4% and earns a net of 5 delegates
Rhode Island: Clinton wins by 10% and earns a net of 3 delegates
Texas: Obama wins by a net of 8% and earns a net of 15 delegates including those taken from the caucus portion of the contests
Vermont: Obama wins handily and nets 3 delegates.
We can fiddle with the numbers a bit, but winning by an extra percent in Texas is worth more than winning by an extra percent in Ohio. If Clinton wins by 8 percentage points in Ohio, she picks up a net of about 11 delegates compared to Obama's 15 in Texas. Let's be nice to Clinton and assume that she manages to eek out a win in Texas, giving her 3 extra delegates. For the day, she'd net only 8 delegates under this scenario -- with Texas and Vermont having cancelled each other out.
Moving on to Wyoming, let's assume, generously, that Obama only wins by 55%. He picks up 2 delegates. Then comes Mississippi. Let's assume the split is 60/40, Obama -- he picks up 7 delegates, and so -- since March 4 -- he's back up 1.
Flash forward to Pennsylvania, and let's assume that Hillary Clinton manages to win 60% of the vote in the state. She'll earn 32 extra delegates -- her biggest net gain so far.
-snipped- (LOTS OF NUMBERS IN BETWEEN WITH THE BEST CASE SCENARIO FOR CLINTON)
The last contest is the Puerto Rico caucuses, which takes place on June 7. Let's give Hillary Clinton an 80 to 20 victory there, giving her a net of 33 earned delegates.
So -- under these most rosy of scenarios -- since March 4, she'll have earned 520 delegates to Barack Obama's 461, having reduced his earned delegate total by about 80 -- or -- by about 60 percent -- but he'll still have a lead of approximately 100 delegates in total... and be that much closer to 2025.