DOOM3 Benchmarks Don't Reflect Card Performance

Just had a look at the Benchmarks (yes it's a little old now) and I think people are missing something. See the new Nvidia cards (and excuse me if I don't spend two hours looking for a link) have an optimisation where if a part of the screen is black (i.e #000000, black black) then, rather than rendering the whole environment and colouring it all in, it only renders the parts that aren't black. So if you are in a dark room and the only thing you can see is a panel across the room, than that is all that's being rendered. So say there was a big machine is over to your right in the dark, that won't be rendered and so boost rendering time and framerate. So in an unusually dark (VERY dark) game like Doom3, the new Nvidia cards have an unfair advantage. And that brings me to my point.

Doom3 is an unusually dark game. So this optimisation will probably only be used while playing Doom3. And so, to use Doom3 as a benchmark tool is wrong and you can't trust these graphs for card performance on next gen games.