you definitely shouldn't take stats as gospel.. just need to frame them right... somethings are very clear, others not so much.
the concepts remain relative, though... e.g. if we could exclude data due to playing with various types of injuries, or eventually account for their effects on results etc etc... it's just one more layer of sophistication to account for.
do that over and over for each possible force/factor and you can do some cool things for sure. i'd love to see what's going on in some of the more sophisticated analytics departments around the MLB.. the ones that are further along the learning curve. they will always been ~10-15 years ahead if they try.. as time goes that's less and less of a difference, however.
plus, a bit of art involved... although not of the artistic kind
.... rather, educated speculation.