Test can be made to favor one or the other but without a independent testing or knowing the hardware and software setup. All the test in the world doesn’t do any browsers any service. And we have to remember not everyone has the same hardware or software setup so all the test does is give a hypothetical result not real world on various equipments.
They actually post a link: http://lifehacker.com/5575535/how-we-test-web-browsers for the hardware/software used.
But that is on only one system and not on varied system so that will never truly tell the full picture of how a broswer really works on various system. The control is never really matches correctly with endusers’ experiences and various hardwares with different broswers.
And your point is all tests are meaningless?