That's only because you dismiss the results that you don't like and choose to favour results that use canned benckmarks, fixed timedemos and cutscenes, you might as well just run 3dmark06 and buy cards based on that alone.
How about this another new game, DX10, whats wrong with the X2?
http://www.driverheaven.net/reviews.php?reviewid=544&pageid=18
Oh gnoes the dreaded rolleyes.

I love it when people like you post and have no idea of my card history and basically hint at me being a fanboy. Nothing short of amusing but i'll leave you to your little dillusions.
As said no review is perfect and using actual gameplay does bring in the element of human error into the scenario especially when they say the gameplay tests are identical.