Timing is indeed unpredictable. Calculations only show (approximately) how a device would behave if made, but cannot tell anything about when some device along the same lines might actually exist -- or whether it will even be worth making, since better technologies often surpass good ones before the latter can be developed.

One application of this kind of knowledge is to suggest what might be worth developing. As Alan Kay said, "The best way to predict the future is to invent it."

Words and ideas I offer here may be used freely and without attribution.