I do think some bowls offer an opportunity to evaluate the relative strengths of conferences. But I also think the opportunity breaks down when a good team experiences a late-season disappointment, and doesn't care about the game, never mind the name of the bowl.
I can think of two examples where Alabama fell victim to this, and one where we benefitted.
Obvious examples of the first were the Utah and Oklahoma Sugar Bowls after the 2008 and 2013 seasons. In both instances, we were #1 until the previous game, which we lost in tough fashion, and were out of the NC race after leading it for most of the season. We went through the motions in New Orleans, and got our heads handed to us.
We benefitted in the Citrus Bowl after the 2010 season. True, we were healthy for the first time all year, and ready to atone for an unfocused three-loss season. But Michigan State was co-champion of the Big 10. Due to a bizarre tie-breaker process, they rightly felt jobbed out of a spot in the Rose Bowl. After having Pasadena dangled in their face, they really didn't care about being in Orlando.
I think we were the better team anyway, but we weren't 49-7 better.
Then you can have times when an historically good team is playing a marginal opponent in a lower-tier bowl in a questionable city. There's a great example this year: Florida plays East Carolina in the Birmingham Bowl in, you guessed it, Birmingham, Alabama. I can't imagine the Florida players giving a rip about that game, or many fans making the trip. I don't know what the bookies say, but I wouldn't be surprised to see ECU win straight up.
Point being, a lot of variables come into play, and the sample sizes are sometimes small. I think you get the best measure when you have teams with good, but not championship-caliber, records (2-4 losses) playing.