I don't engage in such bash threads. Heck, I don't engage in hardly any threads, yet I do read a lot of them. Here's my take on this.
NVIDIA deserves to be taken to the woodshed for this one. Fudging a few numbers here and there is one thing, but the extent that they cheated in a highly popular benchmark is downright criminal. Granted, my world doesn't revolve around 3dMark. (I've only ran 3dMark2001SE a couple of times to see what it looked like.) However, I do see how it could be regarded as a litmus test for general performance.
It would be like GM sneaking a highly tuned Corvette into a one on one comparo vs a Dodge Viper. Then not telling Car and Driver about the tweaks and letting them draw some bogus conclusions.
One also has to take these matters more seriously when considering the ever increasing cost of this hardware. Nowadays, a highend video card purchase must last a good two years to be a worthwhile purchase (for me.) If deception (of this magnitude) is the basis that I arrive at my decision - I would feel cheated.
Two wrongs don't make a right. I don't care what the other companies do. NVIDIA used to produce the best video cards - no questions asked. Now they've stooped to this to somehow try and make up for the SNAFU that was the GFfx 5800 Ultra? Even then, it's not so much the cheating, but the way things were handled when caught. Shame on you NVIDIA.