I agree with a lot of what Awsi saying but I think he's overstating Romney's chances. I've been reading Nate Silver's writing since he was posting as "poblano" on DailyKos, and I thought his early projection on Romney was not very well considered... building hard statistical formulas without enough data for a truly realistic study and applying them to a situation too fluid to be analyzed that way in either case.
A "foundational evidence" argument is much better made on the relatively short time the Democrats have been in power and the dramatic ideological battle being fought within the Republican party. There's no good precedent for Obama's defeat on those grounds. The Carter parallel, often trotted out, ignores that Carter never should have been elected in the first place. The country was reeling from Watergate and Carter seemed the perfect antidote: a pious outsider. Even then, the country slowly realized what it was about to get in to and Ford nearly came back to win anyway despite being like 20 points behind in the polls and facing a stiff primary challenge from Ronald Reagan. The shift in the voter preferences of SAMs (which is a great term) from Democrat to Republican kicked off by Civil Rights was just hitting it's stride, setting the table for what should have been 20 years of avalances were it not for the immense corruption and stupidity of Richard Nixon. It took years for the Democrats to truly reassemble the base from the shock of it. Clinton patched together an awkward centrist framework and combative style that won him elections but made it harder for everyone else in the party. Only in 2008 was Humpty Dumpty put back together again, some 44 years after LBJ's reelection.
I maintain as I have all along that this race is essentially a repeat of 2004, even down to some interesting, almost nonsensical details, like the uninspiring upper crust challenger running a bad race (in many of the same ways) and an essentially invented wedge issue pushed to prominence by the incumbent (gay rights in 2004... reproductive rights this time) to help secure a healthy turnout by the base.
I've seen nothing to knock me off my months old projection of 303 electoral votes for Obama, with Florida the next in line to fall if things break favorably late (which they seem to be doing) and either Colorado or Ohio being the first to fall back in Romney's favor should things go that way. The issues in the margins (of which the economy is the key one) have made Obama vulnerable but Romney has been too weak to capitalize on them. When it's all said and done, that will the story written about 2012.