You correctly point out that a top hitter only affects 11% of the team at bats, whereas the pitcher affects every batter faced.

I counter that by noting that the pitcher plays in about 20% of his team's games, whereas the hitter plays in 95-100% of his teams games.

I think the key to the argument is what does the replacement look like.

If I'm looking at a choice of (Alex Rodriguez + Matt Morris) vs. (Tony Batista + Mark Prior) I think I'd rather have the Rodriguez group.

But if I'm looking at a choice of (Alex Rodriguez + Adam Eaton) vs. (Scott Rolen + Mark Prior) then I'm probably gonna look at the Prior group a little harder.

Reframing the discussion a bit....

The basis for most of my judgements is my belief that year-in year-out hitters are generally easier to predict than pitchers.

I have a few pitchers I have acquired that I trust to carry the load every year. If I look at the team that I feel is my best I have Wood, Morris, Colon, Glavine, Clemens. Many of these guys may not be found on an arbitrary list of "TOP 10 STARTERS" but in that group I can always find at least 3 or 4 solid starters and in any given year 2 of them will probably have GOOD (but not GREAT) years. I've had good success with this team with "SOLID" starters every year and a starting lineup that usually has 7-8 hitters that would come close to all-star quality (I think the worst hitter in my lineup this year is going to be Miguel Tejada). This year I feel a little vulnerable because my bullpen is weak, but I think I can easily fix that in the FA draft.

With a few exceptions I would rather invest my resources in "good pitchers and great hitters" rather than "great pitchers and good hitters" .... My experience is that this approach works best for me.