I watched the debate last night. I thought it was pretty informative.
As I saw it, Obama's main attack on Romney was his refusal to outline his policies in anything other than vague, broad-stroke measures. The problem is that that criticism isn't necessarily conducive to televised debates (where any discussion of policy specifics or hard data is dismissed as esoteric). It also gives Romney the opportunity to pivot back to the centre, which he did very effectively I thought (including on his healthcare proposal, which wasn't the millstone some feared).
In terms of polling, the CNN instant poll gave Romney a 38 point win - the biggest debate win since Clinton mopped the floor with George HW Bush (slightly bigger than Obama's victory over McCain in 2008). In translating debate ''instant-polls'' to actual polls (i.e. how much of a bounce the winner gets), there is a positive correlation since the beginning of 1970, but it isn't statistically significant (isn't all that far off significance, mind), so history suggests any "bounce" should be treated with caution.
In this election, though, I would say this debate was crucial. I think there was a sense over the last month that things were slipping away from Romney; if he had had a particularly bad debate, it could have turned into a knock-out blow. Instead, it will have reinvigorated his campaign and, just as importantly, his big-hitting donors.
I've thought for a while now that if Romney wins, he will do it relatively comfortably; that is, the dynamics of the race will shift against Obama. If it comes down to a close race decided by the electoral map, it is hard to see him winning. Romney has two major demographic problems: women and rural Midwesterners. In terms of the outcome, the latter are arguably more significant. Ohio, more so than ever perhaps, is the most important state on the map. With Michigan and Pennsylvania largely written off, if Romney loses Ohio he is reliant on a hugely unlikely sweep of Florida, Colorado, North Carolina, Virginia and Iowa - basically all of the other swing states.
Republicans typically win Ohio (which typically leans Republican by about 1.8% but is leaning Democratic by about 1% this time) by running up huge tallies in rural counties in the Southern part of the state, offsetting the Democratic vote in urban areas (Cleveland especially). The problem for Romney is that these voters are more comparable to traditional Southerners (many of them live on or near the Kentucky border) than anything else, and Romney hasn't been able to connect with them in a way that someone like a George Bush was. That problem is accentuated by the fact that he opposed Obama's bail-out for the auto industry, which saved a fair few jobs in Ohio, and accounts for polls in the state suggesting Romney faces a major up-hill battle to win it.
Will be an interesting month anyway.
Edit: amidst all the other stuff about winning and losing the debate, I thought we saw the most important indication yet as to what kind of president Romney would actually be in terms of his decision-making process. It is worth watching from about 1.06 to 1.09 here:
http://www.youtube.com/watch?v=dkrwUU_YApE&feature=g-logo-xit
Wouldn't take a genius to figure out he was involved in management consultancy!