There is a huge problem with these numbers: they are meaningless as a method of comparison.
To derive meaning from stats like these there needs to be a control, or at least commonality in opponents.
Unless all 128 teams played a substantially similar schedule, what is being compared? The only stats that could potentially be compared would be those that rely less on an opponent, like penalty yards.
Looking at the SEC, or any individual conference, might be more meaningful if all OOC games were removed from the stats, boiling down the common opponents to 13 possible common options of which every team plays 8.
We might as well throw the Oakland Raiders and Beech High School on that list and see where they come out. Maybe even a Pop Warner team or two.