Rasmussen:
23-38 [-15]
47-51
The strongly disapprove vs. strongly approve business is irrelevant, unless one of the two somehow exceeds 50%.
I would say that the mid-forties is the danger area, while 50% is fatal.
So if next Nov 1, Obama is at 50%A/46%D and 45% disapprove strongly, you would predict he would lose? I don't agree. Similarly, an incumbent with 44% approval and 43% approve strongly, probably still isn't going to win.
It reflects an intensity gap that coud swing a tied race against him based on turnout. But not much more.