I'm wondering if the book from which this comes has discussed any modeling issues that might be relevant -- that is, is this question testing something they have taught, rather than just your intuition? If I were tutoring you in person, I would be looking through this chapter to see what they have said.
The main possibility I see is that a bounce takes a non-zero time to occur, so the "time of a bounce" is not quite well-defined; but if the model is appropriately used to predict height, then it seems appropriate to use it to predict when the height is zero, allowing for some error. I see no more error likely in the times (which are very simple) than in the heights.
If this were a physics text, and explained why the model is used, there might be more to say. It seems to me that two factors would combine to reduce the height of each bounce, namely air resistance (which is a continuous effect) and friction during the bounce (which occurs discretely, and can't be modeled well by the exponential factor, which is continuous). So the model would seem to be paying more attention to the less important effect, and therefore might be bad for any purpose.