I would cap conferences at 12 members and make each P5 team play at least 10 P5 level teams a season.
I would think about saying if the review booth can't make a decision in 60 seconds the ruling stands.
I wonder about the understated byproduct of this would be a dent in the sport's financial fortunes.
One of the interesting quirks in the sport is that quality and popularity are highly intertwined. On average, if your program is more popular, it can create better financial backing and that can buy tools (coaches, facilities, illicit payments. What's also interesting is that success drives finances as well, i.e., if UGA won four games a year it would make less money than if it won nine a year.
And what's interesting about a system that cuts down on P5-G5 or FCS games basically is transferring losses, on average from those worse teams to better ones and more notably from less popular teams to more popular ones. So on an average year, NIU or Western Michigan might go from 5-7 to 7-5, but a mid-tier P5 team might go from 8-4 to 6-6 and be just as good (or a Big Ten team like Illinois might miss a bowl).
Now this factors in because our standards of success likely won't shift much, even if records shift some. Almost one feels better about a good 7-5 than a softer 8-4. So there's a net loss in joy for the most popular teams, which end up drawing the revenue.
(A shorter version would be this, because of the more fish bowl nature of the NFL, more fanbases are not happy with the results than in college. And we'd be moving closer to that sort of thing on the P5 side, which accounts for the most people)
Not that it would be an impediment, but it would be an interesting externality.