Premise of the question may be a tad shaky. The selective use of stats can prove any point.
Two examples of the misuse of statistics:
1. Player #1 (Thomas Jones) had the following numbers in 4 or 5 categories. Player #2 (Arrington) had simililar numbers in those categories. Ergo the two players are virtually identical. This ignores other NFL players who may also have similar stats in those categories. (In other words, if Barry Sanders or Marshall Faulk had similar numbers, wouldn't that mean that JJ is as good as Barry or Marshall?)
2. Who says the statistical categories that were cherry-picked are the end-all? What about other categories as Total Yardage, Yards Per Carry, Percentage of Carries Picking Up 3 or more yards, (Conversely, the Percentage of Carries for Losses), TD's, Carries Per Game, Durability, Percentage of Gains Over 10, 15, 25 yards etc.?
What about the intangibles? How do you measure vision, decisiveness in the hole, blitz pickup skill, faking without the ball etc.?