For now, it's a question that can't be answered, for obvious reasons. But, we do have facts on the poll.
The current dKos/R2K poll has Obama up 51-40. That poll has four regions and defined which state is in which region.
But, if we take the polls on Electoral-vote, balance them by the percentage of the vote from that state in 2004, then the margin is narrower. (I added a few new polls mentioned on Politics1 too). And the total is 47.9-46 for Obama.
Back to regions, we find that the split on the Research poll was 22 NE/29 S/28 MW/21 W.
The split in the 2004 vote was 22 NE/24 S/34 MW/19 W.
So basically the South is oversampled and the Midwest is undersampled. Which means that Obama's lead may or may not be larger.
More number crunching now?
Ok then. Since you insist.
The E-V.com splits by region
NE: 54/40 Obama (compared to 63/29 O on R2K)
S: 50/44 McCain (compared to 53/40 JMC on R2K)
MW: 48/46 McCain (compared to 54/39 O on R2K)
W: 50/44 Obama (compared to 52/37 O on R2K)
Of course, quirks do exist. Every state had a poll in September, but some of the older polls (IOW, over/almost over 2 weeks old) are DC, ID, IL, MS, NE, NM (!!), ND, RI, TX, and UT. So there are some polls with mold on them.
But Right now, We're probably up. But to ripoff Nick Saban, leading on October 1st doesn't mean a lot doesn't mean much without a win.
So. Any thoughts on this self-review of the Big Orange Poll?