In terms of categories, I do think that it is kind of absurd to say 'this stat is good' and 'that stat is bad' but I think that mathematically, in a (H2H...) fantasy season, that having more categories does have a statistical benefit of diminishing the value of any particular stat. By limiting oneself to 6x6 or 7x7, the tendency is to 'cherry pick' different stats, eg holds, to create value for a type of player, eg MR, over looked by the traditional 5x5 scoring.
By limiting yourself to adding one or two categories, I think that it overvalues holds a bit too much and weights things MORE in favor of MR which then, mathematically anyway, devalues a guy like Bonds who can hit a couple of HR/ game, get 10 walks a week and then you end up w/ the Bonds owner winning avg, HR, RBI and the Brendan Donnelly owner winning ERA and H. So the Bonds owner only 'wins' one category outright and only gets a slight edge out of Barry for that particular week. If you expand the stat base to a few other categories, you can enhance the value of a guy like Bonds who is a statistical monster but, at the same time, have some other cats to allow guys like Donnelly to have some value (if they don't catch any liners off their nose this year...)
We use: AB, R, H, HR, RBI, SB, BB, K, E, AVG, OBP, SLG, OPS for hitters and W, L, CG, SHO, SV, HR, BB, K, HLD, ERA, WHIP, K/9 for pitchers and have had different managers in contention both years w/ the same group of coworkers. Sure, some of the stats are goofy and Livan Hernandez led the majors w/ 9 CG last year but it livens things up and seems to encourage a lot of lineup tweaking and ongoing owner involvement. Which are kind of the goals that we have w/ our group. Sure, some people are more into it than others *cough cough*
but everyone seems to agree that it works. So far anyway.
I do agree w/ bleach168 though that 'factoring' RP wins is an almost impossible task, more like a lottery than a scouting exercise. Perhaps there should be an additional stat...VW for Vulture Wins?