it's true you're right, sometimes the tests have no meaning but the initial test with Doom is something else and I do not agree with you it is not laughable absolutely
it remains my opinion and each one his opinion
best regards bro
Well, think of it like reviewing a sports car, but only testing 0-30mph. It IS useful, if you only play Doom, at those settings. But if Anandtech reviewed a new GPU, say Navi, and the only test was Doom, we'd probably laugh at it.
It was chosen because it presents a case use where there is no reason to pay extra for a pricey Gsync display (or for that matter, the higher scale Freesync displays).
The Freesync display selected has an effective range of 48-100hz with LFC. It does pretty poorly at sub 50hz scenarios (eg; dips with AAA titles). The Gysnc display selected has a range of 30-100hz, and it will do so perfectly unless you hit the 20s, which will bring stutter and duped frames. Below 30hz is pretty terrible though.
Again, the test is EXTREMELY narrow. Thus it is useful if a buyer is expecting to always be in the effective range for the optimal performance, but will not get the same results should they be pushing AAA titles at the 3440x1440 ultra settings in many cases with common Freesync GPUs in use (RX470 through Vega64 and now Vega 7).
They also enabled Vsync on the 1080+Gysnc test, which adds lag, a crippling and worse than needless choice obviously intended to further present the best possible situation for this test to favor Freesync+AMD.
Overall it was a bizarre and limited scenario that was dictated by AMD marketing, instead of a wider unbiased analysis. I would trust an Nvidia-dictated comparison equally, which is zero.
To see what I mean about having Vsync turned on, check out this test, which also has some interesting info on how Freesync on Nvidia is still not exactly as good as Freesync on AMD, though that implementation may improve with time.
ran into trouble linking from mobile, look up battle(non)sense test listing unexpected results.