It would be interesting to see ballistics tests on simulated human torsos, complete with accurate skeletal structure, organs and flesh in the proper locations with accurate variations in tissue density. Perform a study over the course of a year during which we would have randomly selected LEOs of average marksmanship unload thousands of rounds of the most widely selected projectiles from each service caliber, from various distances, various lighting levels, and all weather conditions, into hundreds of these simulated torsos, and publish the results. But, that still probably wouldn't be conclusive. My thinking is that you can never account for every variable.
If I try to justify my own caliber choice over another, I can't. I've tried, but I simply cannot say it's better or worse than any other. I have no real-world experience with terminal ballistics on a human body, and have very limited experience as a shooter anyway. I just standardize on what I've had from the beginning, and if it gets too expensive, or there's a lack of new offerings for it, I'll switch. I just hope I can live the rest of my life never having to employ it against another human being.