I seem to recall the past two years those wonderful Pom rankings had me predicted to finish at the bottom and both years I challenged for promotion. Think I finished 4th both times. I much prefer the site I use. Seems to be much less volatile in its projections. I will put out my own projections later this week based on my site and we can see how it goes.
Nothing is going to be perfect, but Pomeroy seems to be the industry standard for these stats. Seems like you and a handful of others use TeamRankings.com, and they may be different by a game or two here or there, but mostly in line.
If there were only one way to look at these things, it wouldn't be any fun, but this does provide a baseline, as would the TeamRankings data.
My strategy is just to look at their records to date, who have they beat, who have they lost to, who has players returning and who plays in weak conferences. So I avoided the big 12, the big east ( the conference isn't great but everyone seems fairly close) and I was going to avoid the acc but figured I would take a chance on Boston college late.
But I am projected to finish last lol. So we will see how it works out. I think I was projected close to the bottom last year and was able to be promoted.
I actually use TeamRankings as my site as you speculated, but then I have a spreadsheet with all the data in it and use a formula to determine the most valued team based on how many wins over the average team in their conference they are. That gives me a good gauge of whether I am drafting teams with a higher ceiling then the average team in their conference or if everyone is muddled up and the difference in wins between the best team in a conference and the median team is only a couple wins then I know I can wait on that conference.
Based on my "cheat sheet" I secured the top teams I wanted in Iona (projected 1 win better then next best at large team) and Oklahoma State (projected 6 wins better then the median Big 12 team I think).