Are we (still) undercounting fielding?

First, this bit from an interview with C.J. Wilson:

    The moral of the story is that a guy like me or Feldman or whatever, who was a reliever, that wants to be a starter for Texas, that should be just an overall positive thing that the organization has come a long way (from years ago) when we had such a stigma attached to being a pitcher in Texas. Now, it's like people are really excited to play here, to pitch with Elvis Andrus at shortstop, and Mike Young at third base, and Kinsler at second and Chris Davis at first. For me, that's a big thing for us, is that our defense is so much better that people are excited to be pitchers now for us.

Taking that a step further, David Pinto writes: "Better defense means less frustration for the pitcher and less work as he doesn’t need to get four outs in an inning."

This leads to an obvious question: Are we undercounting the impact of fielding?

When we "count" fielding, what do we end up with? Essentially, we have plays made leading to runs -- actually, it's usually parts of runs -- saved. Run down that line drive in the gap, and you've just saved 0.23 runs (or whatever). Which is both interesting and useful to know. But you've also saved the pitcher 3.2 pitches (or whatever). Do those saved pitches make our pitcher more effective later in the inning? Later in the game? Do those saved pitches allow our manager to replace our starter in the seventh inning with our third-best reliever rather than in the sixth inning with our fourth-best reliever.

Brainy sorts have made a great deal of progress in separating pitching and fielding, but it occurs to me that there might still be a great deal of work to be done.