1) It respects the idea of outs as a scarce commodity. Two hitters who are both .300 but one of them has an OBP of .400 and the other has .300 are very different values to their teams.
2) To the extent that a hit is more valuable and accomplishes something, HR, RBI, double, whatever, it gets picked up in other stats (HR, RBI, R, OPS).
The argument for AVG over OBP would be that AVG is more of a classic stat. .406 is still a magic number. Also, it's more discussed and generally rewarded. But, these are all, "because they've always been that way" kind of reasons, not that, "this is a superior stat" kind of reason. So if I had to choose one, it would be OBP.
I run one league, though, that counts both of them. We were going to use OBP only but had more pitching categories that we wanted to use (6x6) and cycled through a number of different hitting ones before deciding to use both. Which I think works well.
I agree with the above. Our league also uses both OBP and Avg and it has worked well. Someone like Adam Dunn, who gets on base quite frequently, is undervalued in 5x5 because his walks and overall value to his team are disregarded.
Generally those who have high averages have quite high OBPs. However, those who have high OBPs don't necessarily have high averages. Guys like Beltran, Giambi, Thome, Burrell, Dunn, who normally have sub .300 averages, usually have a high OBP.
To be really broad with it, I would say that OBP is more of a slugger's statistic and AVG is more of a line-drive hitter's statistic. Just because the guys that clear the bases get walked more often, sometimes intentionally, more often half-intentionally where the pitcher is nibbling at the corners, preferring to give a pass than a ding-dong.
OBP also penalizes the free-swingers like Francouer who may have an ok average but an identical OBP.
i like obp more than avg, but i like ops the best and that's what we use. just like a walk should be rewarded as well as a hit, a double should be rewarded more than a single.
as for how it affects players, matthis pretty much explained it. obp/ops is more of a slugger's stat.
i'd say generally, speedsters/leadoffs like pierre, pods, reyes have low obp. i know their avgs are low, but even if it was at .290+, their obp would probably max out at .350 at best, which is just ok.
the only exception (that i can think of off the top of my head) to the speedsters/leadoffs = low obp rule, is ichiro, because his avg is always so consistently high.
mostly, sluggers walk more than singles hitters, probably due to pitchers not wanting to give up bombs. if ichiro/figgins/whoever manages to tag a pitcher, it's a single or double at best. you make a mistake like that to papi, it's gone.
however, there are exceptions to the sluggers = more obp rule as well, as some sluggers are just too impatient and in love with swinging for the fences. guys like vlad, mcab (except this year), soriano (who happens to be a leadoff as well), holliday walk very little, so their obp is low IN COMPARISON to their avg. but since their avg is usually quite high (except soriano), their obp usually turns out fine.
the only exception off the top of my head that i can come up with for a hitter who is not a slugger but still a huge obp machine is brian giles. target him late in obp leagues, as he's probably going to be underrated by most people who are used to 5x5s.