A few points, questions and thoughts:
First, for those of us in leagues that are based on "paying" a round to keep a player, it would fit to have the number of people per grade = the number of picks in that round. This is because the expectation that one could redraft that player in the round would still be there. From your form, there are 7 "A" players. This would correllate to the number of picks in the first round...if you have one of these players, (and they were drafted in the first round), you should keep them with your first round pick. However, the converse of this claims that if you have a player such as Addai to be kept with your first round pick, you should let him go...because he is an "A-" talent, or a second round pick. This fails though, because a player such as Addai would most likely be the first player off the board in a keeper league such as this.
To explain, if the first round last year, in a ten team league, progressed as LT, LJ, Gore, Addai, SA, Westbrook, Jackson, Parker, Bush and Manning, then the A rating asserts that only two of those players (LT & Westbrook) should be kept in the first round. However, this is not the best play, especially if you have a later first round pick. If you picked Addai in the first round last year, then lost in your league Super Bowl, you must keep him. Addai will be off the board by the time your pick would come and you would receive a lesser player in exchange. I would change the grading ranking system to a by-round ranking system. This should be able to be done through the use of projections for 2008 (I realize rankings are not done yet and am in no way disparaging your work. I think it is great that you are providing more keeper content). In fact, the most honest assesment would be done by pick...if you have pick 1.9, these 8 players offer greater value than pick 1.9. Thus, if you have one of these players (or these 8 are being kept), then you should keep your player at 1.9, but if not, then you should free up pick 1.9.
Secondly, if you do work on this, I know there are a few of us 2-QB fools. It would be nice to see FBG go a little deeper in the QB ranks. (Sadly, guys like Kellen Clemens, Cleo Lemon and David Carr all get started in these leagues). Moreover, it may be worth more to keep a player such as Trent Edwards in the last round of the draft than it is to keep Jeff Garcia in round 8. It is difficult to determine this if Edwards isn't ranked.
Lastly, it would be really awesome to link this through the pick value calculator. That could give a definable value as to how much more (or less) value a player who is kept at pick 2.5, but valued at pick 1.5 has over a player kept at 14.6, but is valued at 5.9.
Great work Mr. P, a fun read and I think this could be a real resource for people in 3-8 man keeper leagues.
Edit: There would be an additional piece to taking this to a keeper league that I was thinking about. The pick calculator would have to be used to project not only your keepers, but the keepers on everyone else's team as well. That way, you identify the players who will be kept and by what round. This then goes into projecting who will be available at your pick. If a player has the best player in the draft, but is keeping the player in the fourth round, then the draft order will have to be adjusted. Instead of pick 1.9 being worth the ninth best player, it will actually be worth the tenth best player. This would be extended for every player that is kept that is better than your pick. If the top thirty players are kept, then the correct play may be to keep the thirty first player available ahead of where he is projected to be kept, even if this is at a pick above the projected value of the player. Simply saying that a player is a specific value may be a bit of over simplification.