I think changing the way we coach pitchers based on this information is way off in the future, if we figure it out at all. However, I think this information will very shortly become a major factor in recruiting and scouting. As more and more pitchers draw up to the 94, 95, 96mph mark, recruiters and scouts will begin looking for qualifying factors to separate high velocity pitchers and you may even see a revolution equivalent to the Money Ball period where colleges and mlb clubs with limited resources try to find untapped scouting factors as they dig through the bargain bin for good deals and overlooked talent.
I think use of this information for "coaching" is not too far off. Folks looking at this data, as a starting point, can probably determine what pitches should be most effective given how an individual throws. What may be a little further out is training/development based on the data, but I could see some near term benefits at relatively young ages. Put this data together with some super-slo-motion and then move on to computer aided "overlapping" (not sure what the actual term is but think 10 images overlapped - I've seen this in Olympic downhill skiing) and folks can sit down and see what body movements are creating a particular pitch as well as see how small differences in motion (not really evident to the naked eye) affect a pitch. Once cost comes down, I think it will grow rapidly.