I was on-board with it until I realized how they were doing the point-differential step. It sounded good at first, but knowing our experience in 2008 sheds light onto the problem. By ONLY using the point-differential to eliminate one team and using head to head on the remaining two teams you are penalizing the team that won by the greatest margin. That team will always be eliminated in a 3-way tie (if they all have only one loss in conference).
If you are going to use point-differential as a tie-breaker, you need to use the point differential from ALL the head-to-head games (or even better, the total point-differential from all the common opponent games).
If you did that in 2008, you’d have the following results which are very enlightening:
OU = +34 (+44, -10)
UT = +4 (+10, -6)
TT = -38 (+6, -44)
If the current system was in place in 2008 you would have had a weird situation where OU would have been trying to let TT score to keep the game close. That’s absurd!