What polling are you talking about that shows that Sununu has pulled even or ahead?  The closest he's come recently was being down by 4 in a UNH poll and down by 6 in a Rasmussen poll.  An ARG poll taken between those two had him down 22 (not that I'd put a whole lot of stock in ARG).

My own theory, based on the exit polls from two of the elections already held, is that down ticket Democrats are really benefitting from increased voter turnout among majority black districts. This factor would mean little in NH.

As for the referendum, giving Obama the spotlight gives him a chance to tell the narrative he wants regarding Bush (easy story) and how McCain is Bush III. Otherwise, the biggest hurdle is setting the bar too high during the referendum, whereby no matter what Obama says it won't be enough.

Eve,

It's a consensus that if the election is about Obama only, then it'll be harder for him to win.  On the other hand, the ballot will have a number of names on it, the top two being Obama and McCain.  So, instead, the election is actually about choosing from the people on the ballot and picking the one you would prefer to be president.

I know there are plenty of groups - the media, the Republicans - who want this to just be about Obama, but people, including you I guess, need to be reminded that presenting the falsehood that it's only about him is a lie that plays into McCain's hands.

I know it's not all that important to the post, but I just want to point out that the term "statistical dead heat" is rather stupid as applied here.  Shaheen did not have a significant lead in the UNH poll (Margin of error was 4.5 and the lead was 4).  That means that we cannot say with 95% confidence that Shaheen is leading based on that poll.  We can, however, say that with about 92% confidence (if my math is correct); "dead heat" of any sort is not an accurate description.

Like I said, I know that the fact that it shows the race tightening is the main point here, but it's just an aspect of reporting on polls that I find a little annoying.

"But we tend to assume the immense interest in Obama, specifically, only works as a positive"

What do you mean 'we,' paleface?

AlanSP's post reminds me that political writers really need to take a refresher class in statistics. In general, I'm not picking on Eve here, but very few if any of the poll-obsessed writers seem to understand uncertainties. On the one hand it is understandable, because a many things about probabilities are non-intuitive, but at some point professionalism ought to kick in. Just my opinion.

I think GSpinks' theory is mostly right, and I am sure Obama's people are working hard to register black voters, and I think they will turn out in huge numbers at the polls. In several states, that will be a big deal and my guess is that people who have never voted before or rarely vote are more likely to vote straight ticket.

Obama has also energized some constituencies that rarely vote, and only time will tell if they actually will vote. That could bring a lot of downticket votes as well. And if not, then I think some of Obama's negatives will not hurt Democrats very much. Although a distressing number of voters will not vote for him because of his skin color (even if they say it is really because they think he might be a secret mooslim), I think only a few would vote against downticket Democrats just because they aren't willing to vote for a black guy for President).

"from increased voter turnout among majority black districts"

Correction: that should have been "increased voter turnout among blacks in these districts".