He may be right in this instance but we can't take Linus at face value. I get the impression he is heavily sponsored by Intel.
I'd just like to see reviewers try to show both sets of hardware in their best light. I don't care what settings are used just get the best results from each and present the figures fairly.
I think the time is coming where consumers will have to subscribe to reviewers and no sponsorship or freebies will be allowed.
Agreed on the latter for sure.
Setup patreon.
Resell hardware when done with reviews to recover bulk of purchase cost, patreon funding should cover rest along with ad revenue.
I would like to see different reviewers each have their own testing methodology.
One reviewer all hardware in best light, max OC, best ram, tuning etc.
Other reviewer OOB testing.
Also when games are tested instead of say 10 reviewers testing the same 8 games between them have them each use different games so we have a far wider range of games tested. I feel games tested should include at least 2 games from each of these categories.
Games not supporting more than 1-2 threads. (games that thrive on per core performance)
Games supporting 4-8 threads but low utilisation.
Games supporting many threads and heavy utilisation.
Triple AAA games
Indie Games.
Vulkan games
DX9 games
DX11 games
DX12 games
At least 3 hardware generations need to be shown on hardware comparison graphs.
Measure other metrics not just basic frame rate performance.
Power consumption during the game.
Stutters, min frame times when locked at 60fps (probably most common config), I have seen games have low cpu utilisation and gpu utilisation so in theory its a cakewalk for the hardware but then is a stutter fest in some configs.
Loading times during games.
Temperatures during games.
Also when comparing say a gigabyte 1080ti to an asus 1080ti, there is no need to redo performance tests, as thats largely based on silicon lottery, just do temps test in those situations.