By far the hardest part of collecting all that GPS and accelerometer data is figuring out what to do with it. If I were going to start making data driven decisions on when to prevent injuries, I'd first seek to predict injuries. That means collecting data for years and using that historical data to tweak my predictive models. The more data you've got, the more resolution you'll have in your models, and the greater the benefit. What I'm saying is that if there is anything to the data science of injuries, I really hope we've got some historical data built up to work with.