Not to pour cold water on this but maybe I am, I am an engineer as well.
I think the brute force approach of placing a couple of pressure / angle sensors / accelerometers will not work for skiing. There are multiple reasons just to list a few:
-- Immediate feedback is not as important as it seems. I view skiing as a mostly prediction activity -- in order to get where you need to be at any moment you have to set up and do so many things in advance way before that. Let's say you messed up the transition and will be in recovery mode for the rest of the turn. Will you be trying to tip more when the turn is ruined already? Being able to tip more in certain phase of the turn is possible because I set it up way before and not by doing one move, but by multiple sequential activities. Similar examples can go on and on.
-- How would you decide on the needed CA / CB or an arms position based on the pressure on the sole or angle of the boot? What if you counter rotate the shoulders but not hips? So are you going to put more sensors around the body? The challenge will be mind boggling -- measuring sensors output is not hard, coordinating all this information and most importantly providing meaningful advice based on it seem not so easy.
-- Watch multiple robots skiing on youtube. This is the state of the art and it looks ugly.
What if I place tiny sensors on the fingers of both hands and record the information and then for example would try to teach someone to play piano based on such setup? I think you are talking about similar challenge.