OK, thanks! Yeah, Nate Silver also does a daily blog post on TNR about the day's polling, he's a godsend.
To be honest though, you would have to compare Obama's performance among all these groups with how Kerry was polling among these groups in July 2004.
I have no idea where to find such crosstabs from this time in 2004. But at this point in the race, late July early August, John Kerry
was ahead by an average of 2.5-3% in the polls. Of course, by the time the exit polls were published on election night, he was behind by some 3.5%.
So between this point in the race and election day, Kerry's margin vis-a-vis Bush went down some 6%. Looking back at Nate's table and adding 6 points to Kerry's margin, hmm, you get numbers very similar to Obama's now. Which makes sense, since Kerry was up 2-3% in the polls then and Obama is up some 3% now.
What to take away from that? I dunno. There's a lot of structural reasons to believe the climate for a Dem is better now than it was in '04. But just looking at the polls, Obama isnt doing much better right now than Kerry was at this point in time. So it's close.
(One note of comfort: Kerry's 2.5-3% lead at this time in 2004 was about the largest he got to enjoy in the whole campaign. He'd been doing clearly less well throughout the year before July. Obama's lead has in comparison been more robust through a longer period of time, which hopefully signals that it will also not deflate like Kerry's did.)