FWIW, I'll toss my opinion in here. Until I see more, I don't think you'd be far off if you just took OU, Bama, Mizz, LSU, Texas & PSU and drew their names out of a hat for your top 6. We just haven't seen enough real tests yet. As The Bear said, "You never know how a horse will pull till you hitch him to a plow." Of the top 6, Bama's been tested the most and looked like world beaters at times, but mediocre at others. Can't seem to string two complete halves together or finish opponents off. Saban has good players playing like great players though, and that's worth a lot. Talent on our OL is exceptional but there's almost no depth. An injury there could be disastrous for us. LSU has enormous talent but we haven't seen a real test yet. The Auburn game was a great win for you, but they're now looking so bad that I don't know how much that game really showed us about you. I expect you'll beat UF this week and we'll have a better idea. As for the other 4 teams, they simply haven't really played anyone yet, but that's soon to change and we'll see who's for real. All 6 of us have some major strengths but a few questions remain to be answered. I'd say Bama & LSU are as good as anybody else in the country right now, though.
I'm not one for preseason polls and I do agree that polls shouldn't come out till at least a few games have been played. However, we started the season with a HUGE question mark at QB. Nobody in their right mind would put a team #1 with such uncertainty at the most important position on offense. Since we do have preseason polls, somebody had to be number one and UGA was as good a candidate as any. Of course, now we know they aren't and therefor they have dropped.
In creating a computer rating system, I found the most accurate method of rating teams early in the season was by simply starting where last season left off. Therefore, LSU should have started as #1. Speculation ain't worth much.
You take where they left off last season, factor in what they lost from that team and also whether a team was young that year (more than likely making them better following year) and then come up with an educated ranking. It's a bit ignorant and uninformed to simply start each team where they left off the year before when there are so many factors involved.