I acknowledge that's the reasoning they might have used, but it's shaky. And, I'll use it more as a basis for discussion on what it means if that's how they make decisions, generally.
We're not a two gapping team so playing the "run stuffing/contain role" has some, but lesser value to what we do. Guys have to make plays. On any given play, in any given game, we need guys who can make a play based on individual ability more than on assignment. Think the opposite of the NE Patriots.
Rousseau has shown he can do both and isn't just a wild upfield player. Their mistake was giving Patchan too many snaps using the "let's stuff the run and play contain against UF" logic - as if Patchan was necessarily superior once things broke down inside of a game. I'm not saying Rousseau should have been the outright starter in Game 1 (though maybe he should have been given he's pretty much always looked like this in practices we've watched), but the snap count was Garvin 49, Patchan 43, Hill 15, Rouseau 14.
Those numbers, from a staff truly focused on analytics, might have looked better as Patchan 28, Rousseau 24, Hill 20. Anyone watching yesterday saw Rousseau extend his arms, accelerate off his block, and make a run stop on the edge. A play I'm not sure anyone else on the team can make. He can make plays with his wingspan and speed others cannot. Those extra plays against UF might have made a difference, who knows.
As you noted, to double down on that versus UNC was a travesty that plays a role in keeping us from winning the Coastal. Against UNC, Garvin received 59 snaps, Patchan 45, Hill 24, Rousseau 12. Let's call it what it is: a shaky personnel decision we didn't adjust on quickly enough.
I'm sure there are halfway decent rationalizations and "he wasn't ready" or "we didn't see consistency in practice" (even though you and I know anyone who's seen practice saw nothing but absurdities from Rousseau) or "we wanted to ease a young guy in his first couple games." It's over now and a healthy Rousseau will be on the field, but unless the methods used to make these types of decisions get adjusted, we'll just find ourselves a step behind in another scenario (like starting two Frosh tackles in Game 1; not emphasizing the pistol/Shotgun sooner, etc.).
We shall see.