We crown champions on the field because that's the way sports works. If we just eliminated all variance and used metrics to crown a champion, sports would be lame. The Patriots would be 2007 Super Bowl Champs.
But I'm not crowning a champion. I want to know which teams and conferences are best because that's what I'm interested in. Lots of people are interested in it, and not in an ESPN talking heads or message board nonsense way. Also, determining the place of teams in college football despite the asymmetrical schedule is a necessary evil because of how we determine which teams get into the playoffs in the first place. For example, if we truly want the four best teams, Florida State does not make the playoff. Florida State does not even make the 8-team playoff. But it's clear that either we're still antiquated in the ways we think about things, or we value coming up on the right side of variance and putting together a deserving resume. Personally, I like building a resume to get in since it keeps more of the qualification on the field, but FSU still does not make a 4-team playoff with their resume if we take margin of victory into account. They do make an 8-team playoff though, resume-speaking. Also, it's entirely possible within the realm of probability that the best team in the nation this year has 2 or 3 losses - not all that likely since the best team is one of either Oregon, Bama, or TCU, but still entirely possible.
I think it's incredibly valid to discuss which team is truly the best team simply from a fan standpoint and someone who is interested in understanding the CFB landscape, while also recognizing that the champion is crowned on the field. Real results have to matter otherwise sports is pointless. I could sit at home with a random number generator and simulate seasons but where's the fun in that? Besides, every metric ever is based on those real results. So, once again, I do not buy the "why play the games when you could just look at your computers" nonsense. The computers are based on the games.
Another thing people don't understand is how widely game results vary. Auburn is a 6-point favorite today. Does that mean that they should win the game by 6 points or by 5-7 points every time? No. They play that game 100 times, Auburn will win 49-10 sometimes, and Wisconsin will win 35-13 sometimes, and they'll tie and go to overtime sometimes. On average, Auburn is predicted to be 6 points better - and this late in the season, those lines are incredibly sharp. That's all it means.
If we care about how conferences stack up and one conference goes like 0-9 or 1-8 or something against the spread and against the computer predictions, that would certainly be meaningful information and would tilt the way that conference stacks up. But a conference going 3-6 or 4-5 against the spread or 5-4 or 6-3...those are all very likely outcomes and shouldn't change the way we think at all. My point is that, and I've already seen it in this thread, people cherry pick the results that fit their narrative and ignore the ones that don't. Personally, I hate all narratives - I don't care what the storyline is. If I flip my view on a team or computer rankings change for a team, that's not flip-flopping or waffling or even being wrong. That's called adjusting based on new information. I work in science. This is what I do.