Revotes in Michigan and/or Florida may not be so good for Clinton
What would happen if an agreement were announced today that there would be re-votes in Florida and Michigan? Immediately, the previous primaries in those states would become dead letters. Instead of being 200,000 votes down in the popular vote (by her campaign's count), or 500,000 down (by my count, which gives Clinton her Florida votes), Clinton would be down in the popular vote by almost 1 million. And 193 delegates that they are currently counting would suddenly disappear.
And at that point, the magnitude of Clinton's deficit would be too obvious to spin away. Yes, there would be two additional large-state contests in which to win back the million popular votes and hundreds of delegates. But unless she did significantly better in both states than she did in the illegal primaries, she would lose, not gain, ground, by her own calculations. Since she was on the ballot alone in Michigan before, it's highly unlikely that she will do better there. It's very possible that she could do better than the 50 percent she won in Florida in January, but since it would now be a two-person race, it's a dead certainty that Obama would do significantly better than the 32 percent he got in January, thus adding to his total popular vote margin and delegate count even if he lost again, and so it would be a net loss for Clinton. Re-votes cannot help Clinton be "perceived" as the winner of the popular vote.
Contrary to the gullible media's belief that "time" is a "powerful ally" on Clinton's side, in fact, Clinton's only ally is uncertainty. The minute it becomes clear what will happen with Michigan and Florida -- re-vote them, refuse to seat them, or split them 50-50 or with half-votes, as some have proposed -- is the minute that Clinton's last "path to the nomination" closes. The only way to keep spin alive is to keep uncertainty alive -- maybe there will be a revote, maybe they'll seat the illegal Michigan/Florida delegations, maybe, maybe, maybe. In the fog of uncertainty, Penn can claim that there is a path to the nomination, but under any possible actual resolution of the uncertainty, there is not.