Its the absence of any sort of scientific method to it that gets me! The hardware is a significant element but you'd get divergent results using different video drivers or graphics API's, or different memory configurations, etc, etc.
The relative power between the hardware won't change.
Showing the variety of games already shows how different software can be more or less performant, there will always be differences there, it could be a great/bad pc port, or great/bad console port. That's a given and it's borne out in the results, you'll never get a complete like for like between console and pc hardware.
The video just shows you will get a pretty similar experience except when you crank up the resolution due to the limitations of the vram.
I'm not sure why you would want to test different iterations of drivers, 99% of people buying graphics cards just update to latest driver as it's generally the most performant for latest releases.
Again, I don't know what you mean by different memory configuration, like slower and faster ram? Single or duel memory set ups?
Why is that important when we want to see the relative performance of the GPU?