I have no clue what their formula is, but if you multiply 6.1 by the number of 6.1 rated players, 6.0 by the number of 6.0 rated players, etc., and then add them up and divide the result by the total number of players, USC has a higher number. This would give you an average rating of the entire class. Again, not sure if this is what they use, but that's the only way I can see them rating USC higher.
That gives them the higher average rating per player. But if you take that theory, a program could sign one 6.1 rated guy, and no one else at all, and have the best class in America (because they would have the highest average).
One thing we do know by comparing the classes, is that typically, the more guys you sign, the more points you accumulate. Take 2007. Alabama finished #1 with an average player rating of 3.72. Notre Dame finished #2 with an average player rating of 3.96. That's a huge difference. But Alabama signed 32 kids, Notre Dame signed 23.
That's one reason Auburn is ranked #4. Our average rated player is 3.5, which is lower than Alabama's 3.62, Oklahoma's 3.55, and even #12 Penn State's 3.55. Thing is, Auburn signed 32 players to Penn State's 20. In doing so, we accumulated more points, even though our "average" recruit was rated lower (although the top of our class was better than the top of PSU's class, those extra 12 guys we signed drag down the overall average...but also add a lot of points).
We can see this fact all up and down the board. There are teams with lower average ratings, but more signees that are ranked higher. There are teams with high average recruits that finished lower because they didn't have many guys (Cal averaged 3.58, but only signed 19 kids...they are #11).
USC seems to be the exception. The top of their class is nearly identical to Florida's, and Florida has more recruits. In every other example, if everything else is the same, "more recruits" equals more points. How is this not true when it comes to USC and Florida?