[noob]

It could have been biased towards AMD, even though those tests are thought pretty reliable.

What I would like to see is several different benchmark tests run for each section.
say not only half life 2, but also maybe battlefield 2, farcry, and a couple of others.
add a few more of the other benchmarking tools - say 3d mark 05.
And others for all of the rest of the sections.

That would've given a better comparision.

You should never show 'unbiased' using only 1 benchmarking tool per section. Thats like saying AMD is better than Intel because 'x' benchmarking tool says so. While 'y' benchmarking tool favours Intel, and 'z' benchmarking tool favours SunSparc and both AMD and Intel run like dogs on it.