I tried a variation on this idea recently, with the following lineup:
C: 1958 Berra
1B: 2007 Youkilis
2B: 1966 Helms
3B: 1954 O'Connell
SS: 1989 HoJo
LF: 1937 Ott
CF: 1946 Hopp
RF: 2001 ManRam
Not a high range team, but each of the above guys fielded 1.000 at the position indicated (which was often not their primary position).
The team finished 86-76, but were in a weak division which they won rather handily, and I rested the starters quite a bit down the stretch...they were probably a 90 win team. The team fielding percentage was .996, with a total of 23 errors.
Of those 23:
1 was made by a bench scrub
3 were made by pitchers
14 were made by AAA players
and 5 were made by the starting lineup...all 5 by Yogi Berra, and I believe all 5 were throwing errors during an SB attempt. Other than that, the lineup fielded perfectly.
The team played in AFCS and scored 874 runs. They gave up almost 1000 but many of those were allowed by AAA and mop ups...the regular pitching staff was quite good:
08 Joss (27-11, 3.12)
09 Summers (23-14, 3.72)
81 Righetti (6-8, 2.51 as the front half of a tandem)
19 Nehf (12-7, 4.38 as the back half of the tandem)
18 Toney (7-1, 2.66)
plus a handful of other guys.
That's more cookie than I like to have on the mound, but the defense was expensive enough that I had to emphasize value with the rest of the roster.
The team lost in the 2nd round of the playoffs.
As with many of the high-range examples noted above, I believe it's possible to put a competitive team in an OL built around fielding. The only real cookie in the lineup is HoJo; the rest of those guys aren't seen all that often. This was only the 2nd time Helms has been used in an OL, according to his performance history, for example.