Would it be enough to 'catch' on something, say an improperly sealed section in the landing gear system, and cause it to heat up or divert the path of the shuttle while it was turning?
Dunno the specifics. However, as the speed of the shuttle is high enough to use the control surfaces even at that altitude, I'd have to also believe that there is enough to snap off pieces like the "fall-off" tiles or a landing gear door. The most-likely effect of that, because tiles do fall off, is to increase heating rather than create an assymetrical drag.
So then it would be a question of 'how vulnerable is the exposed section to heat'? If the tiles do fall off, with no effect to date, then I would assume that some sections can withstand not having them better than others. What sections on the shuttle could cause this type of catestrophic failure due to overheating.
As of now we don't know if the 'explosion' the overheads saw was cause or effect of the disaster, but I imagine that finding out what overheated will lead to the culprit.