I'm not sure whether this would fall under your definition of "reliable" (or indeed, how useful it would be) but Carl Franklin's book British Napoleonic Field Artillery has a list of casualty predictions for given calibres of gun firing at different types of target, over likely battlefield distances/times. I believe the predictions were obtained from various studies, mostly (possibly all) from firing at targets, rather than actual people. I have the book (although not immediately to hand) and could check on the actual sources if you would find it useful.
I believe that Larrey made a study of French casualties at some point in his career (is that what you already have?), but obviously that will only refer to wounds, not to deaths.