Regardless of what kind of bounce Romney gets out of the debate, or how close the national polls get, the electoral college hill he has to climb to reach 270 is virtually insurmountable.
If you take Romney's RealClearPolitics polling averages in all the battleground states and add 1% to all of them, he wins North Carolina and loses every other swing state, which results in a 332-206 electoral vote victory by Obama.
Add 3% to all states instead, and he manages to flip Florida for a final tally of 303-235.
Add 5%, and he's able to flip Iowa, Virginia, and Colorado but STILL loses, with Obama winning by 275-263.
Only at around +6% does he win Nevada, and bring the electoral vote score to a 269-269 tie, and give him a controversial and impossibly narrow victory from the House of Representatives.
In other words, Romney must improve his standing in each and every battleground state by at least 6% in the course of the next few weeks, in order to squeak out one of the narrowest victories in U.S. history. Good luck to him, but my bet is still on Obama...
Actually, Romney just needs to improve his standing in each battleground state by the amount he is behind in
that state, and I see no reason to expect the swing to be uniform. So, given the nature of Ohio and the nature of Obama's lead there (he is only ahead there by 5.5% according to RCP), Romney should be able - in fact is very likely - to turn it around. Even though the polls say Romney is closer in Nevada than in Ohio, I think the actual likelihood of a Romney win in Ohio is much greater, given the nature of the electorate in each state.
The most likely map on election day is this:
