Pollster.com

I'm amazed to see states like Montana and Alaska that have been deep red states for years and years finally coming to a tinge of possible blue for this election under Obama as the nominee.

Alaska can be partially explained by the fact that Stevens and the Republican Party have utterly destroyed the party "brand" there...still that Obama does so much better than Clinton shows some real issues.

The SUSA poll of about a month back also showed that Obama placed more "Red States" (and cross-over States) in competition than Clinton. Still Clinton supports might point to the Ohio results (although statistically insignificant) and say that she could win more big critical States, and does better in the Blue States. However, in a "winner-take-all electoral College you don't get extra points for winning by 65% than by 51%. So the goal is to get competitive and win in battleground states. This also should help the party in the down-ticket races.

In the SUSA poll Obama also did quite well in Texas. Clinton did better in Florida.

One thing these polls should begin to evaluate is how well the different candidates generate "Likely Voters". Earlier polls have often failed to correctly model who would come out and vote from the pool of registered voters. It should be simply to correct their estimates of likely voters from various demographic groups based upon increased turnout in those demographics from what they originally modeled.

They also need to establish some weight for increased Party turnouts based on Primary participation.