It's an interesting way to look at it for sure, but it still feels a bit cheap to me to just throw away all those losses. P5 teams also have repeated matchups against the same teams (that's what conference games are) and can't just ignore a bad matchup or put extra weight on the ones they play less frequently.
I think the various comments have hit pretty well on this though, with some advocating to throw out games against Utah, others pointing out that they shouldn't be. I tend to fall into the latter camp, but I'm happy to at least provide the data so that people who truly believe that is a better reflection can look at it.
Ultimately, I don't think it changes the final conclusions that much though. BYU is either playing (most likely) at the level of a subpar (GaTech) or (best case) mediocre/average P5 (around #35 nationally) over the last decade. Even considering my belief that they've been subpar for a P5 team, it's impressive to do so without all those resources. Utah is, at this point, generally a fixture in the top 25 (present year excluded obviously) and so probably ever so slightly above average as a P5 team after a rough initial acclimation. Given how close the two teams were in the MWC/WAC for the two decades before going Pac12/Indy, I see no reason to believe BYU wouldn't reach similar heights if they ever found a P5 home.