I don't believe that hawks use their prey as collision rebound devices; I've never seen or heard of such a thing. It also sounds like more planning/better built-environment sense than a hawk might have enough brain cells for, especially when doing all the other calculations necessary to take prey on the wing. I'm not sure their eyes are up to it either.
For example, an American Kestrel has nearly twice the visual acuity of a human, but its peak resolution maintains only over a range 55% as wide as a human's. That sounds like a design for exquisite viewing of what's right in front of it's beak, but not such great peripheral vision--which makes sense, given how it makes its living.
Added to all the other vulnerabilities bird's have for crashing into plate glass, the limited peripheral vision of raptors--other kinds of birds may need better peripheral vision to steer clear of them--just increases their chances of crashing. Accipiters are the raptors most likely to hunt around human habitation, especially when attracted into our yards by feeder birds; hence, the most likely to glass-crash.
I'd say that was one very lucky Sharp-shinned and one humiliatingly unlucky dove.