Let's revisit your premise.
Alabama would lose to Texas. I'm not trying to relitigate results, but if I'm anticipating future results like you are, I should use all the available data.
One point of data is that Alabama lost to Vanderbilt. That's true, they lost to Vanderbilt by 5 points. 7 of those points though came as a directly result of a coaching blunder, which had nothing to do with how the team played. Likewise, Texas only beat Vanderbilt by 3 points, so logically there's no basis on which to use the Vanderbilt result as proof Alabama would lose to Texas.
Furthermore, Alabama beat Georgia, which beat Texas, twice! That's a pretty significant data point as well.
Basically the only thing you can hang your hat on is the Oklahoma game, but that game had multiple mitigating factors, including two interceptions one of which caused by a receiver not making a block, and a TD that was completely legit being called back.
Not excusing that loss or pretending it didn't happen, but I see nothing there that proves Alabama loses to Texas either. It was Alabama's worst performance of the year, which would be their floor but not a likely future result either (Notre Dame lost to Northern Illinois but still beat Texas A&M by the same margin as Texas and Indiana soundly).
Texas played a much easier schedule hence they had a better record. They didn't play Tennessee, South Carolina, LSU, or Missouri! It was the easiest SEC schedule of any of the contenders, and we all know teams playing soft schedules look better doing it (see Indiana). Anyway, I'm just not sure what you're basing your argument on other than the entirely outdated wins and losses without analyzing said results method.