Scott, I've got to call baloney-at-first-reading on this claim. If I read you correctly (please correct me if I'm wrong), you're saying that at least 80% of 90-95% of all shots are half-ball hits. So (multiplying 80% by 90% and by 95%) you're saying that 72-76% of all shots are half-ball hits.
I just conducted a non-scientific experiment.
In Trial A, I rolled all 15 balls and the cue ball randomly out on the table, and shot the balls in as if I were playing straight pool, but without attempting to save a break shot; just shoot the 15 balls in the holes. Then I repeated that for another rack, and another rack, until I had made 100 balls. I missed three times (two were on careless shots; one on a difficult shot where I tried hard to make it).
In Trial B, I did the same thing, except I shot every shot as a half-ball shot, even when it clearly was not. I tried to choose the shots as if I were really playing straight pool; i.e., I did not choose the patterns based on trying to make everything into a true half-ball shot. I abandoned Trial B after three racks, because the conclusion was obvious. By the time I had made the 45 balls, I had missed 43 times.
This experiment was certainly not any sort of rigorous proof of the falsity of your claim. But it quickly convinced me, at least, that the percentage of all shots on a pool table that can be made by aiming the center of the cue ball at the edge of the object ball is nothing like your claim of about 75%.
(And, just to be clear, aiming the center of the cue ball at the edge of the object ball is not at all what is done in CTE. Fractional aiming systems and CTE are wholly different.)