Well I never thought of it that way.
I have yet to ever see my 6800GT in a benchmark against other 6800GT's. I thought that any GT against another would be similar in scores given slight difference in memory latencies used.
I wonder know if my system scores high compared to other GT's, but according to futurmark is did well on my 3500 clawhammer with this card. However, I am cheating the score too, turning all setting on the minimum: nvidia high performance setting, no aa no af, vsync off.
6000/6100 clawhamer 3500
13400/1170 clawhammer 3500
5850/6500 Opteron 170
14200/1300 Opteron 170
Aquamark 85300 Opteron 170
this was with the card overclocked 420/1100 and 170/3500 overclocked
2540