You know your candidate is bad when you're counting on economic conditions to deteriorate for him to win.
Just like Ronald Reagan and Bill Clinton.
Considering you're a big follower of finding parallels with the Gallup poll, I'd think you'd know that Reagan and Clinton had double digit leads on their incumbent opponents in the summer of the respective years. Add in that the third party candidate in both cases like would have/did break for the challenger, and you'll see Obama is doing much stronger than Carter or Bush Sr at this point.
I'm actually looking at approval numbers, as opposed to the horse race number.
Which is cherrypicking. If your matching of Gallup numbers holds any sort of weight, it needs to show a relationship with the actual election results. In the case of Obama-Carter and Obama-Bush on approvals, the challenger ran far ahead of the President when approval ratings were at a certain level.
Your hypothesis is that Romney will win on Obama's unpopularity because Reagan and Clinton did. Your hypothesis fails because the data suggests Reagan and Clinton did not simply win on incumbent unpopularity and, in fact, ran far ahead of the disapproval rating, even beating the incumbent during times of positive approval, while Romney just muddles along.