It may be helpful for people that decide to work on this project (which I would love to see get off the ground) is that the variable in the equation is time, in frames, for the opponent to counter-attack. Everything else is a constant that is determined by the characters fighting (both the attack being tested for and the character being tested against).
A major problem i see is that it is not only important to measure the distance that the opponent is sent away from an attack (from a successful hit or even a non-powershield), but also the ANGLE from which they are sent. Thus, an opponent's DI also affects how quickly they can counter-attack. Also, that means that you are going to have to deal with vectors, and that makes your equation more complex than it already is.
For example, let's say G+W turtles (Bair) a med-to-high-% Fox. No DI sends them diagonally, I think. So the most efficient thing to do time wise would be to spend t1 frames fast-falling and then t2 frames using his B move, or perhaps his side B
However, if Fox DIs down-left, the time t1 is reduced (because Fox can reach the ground, where his opponent is in range of the B and side B), even though the TOTAL DISTANCE that Fox is flung is the same. To find the exact amount would require a study of the maximum angle of change that DI can achieve as well as some trigonometry (since we would have to find the angle of DI that minimizes the distance/velocity for the opponent character to be in range of his most efficient counter-attack from the position he was sent)