Eric Kleefield has a post up about recent polls for the Pennsylvania senate race which I think are rather significant. In a Specter-Toomey matchup it's basically a dead heat. A little while back, that same matchup had Specter with a 9% margin with a fair amount of undecideds (which isn't huge but it isn't tiny either).

Meanwhile, Sestak versus Toomey is much closer with Toomey leading Sestak 39% to 35%. Kleefield notes that a big factor in those results is that Toomey and Sestak both could earn considerably more name recognition. So on the nominate-who-could-win argument --which I don't like-- perhaps Sestak would be the right choice.

To make matters worse, in the Specter-Sestak matchup Specter leads Sestak 55% to 23%. Sestak probably doesn't have a lot of name recognition there either but that's still a huge margin to fill.

Over all, I think things don't look good for Sestak or Specter. Sestak seems unlikely to beat Specter in the primary and Specter seems increasingly unlikely to beat Toomey in the general election. Toomey must be pretty happy.

I want to also note that Sestak's name recognition problem is very serious. A few days ago I met a girl from rural Pennsylvania, a Democrat, who hadn't heard of Sestak. She's the exact kind of voter that Sestak needs and that she had never heard of him is a big problem.

UPDATE: Nate Silver writes: Ultimately, I'd still rate either Democrat as about a 2:1 favorite over Toomey, this poll not withstanding. But Pennsylvania will certainly be moving up in next month's Senate race rankings.