Ah yes, I remember that old but good video.jorimt wrote: ↑22 Nov 2020, 12:29https://www.youtube.com/watch?v=E3wTajGZOsA
It's more specific to the flip modes available to DX12 in Windows though.
That said, beyond subjective per-user reporting, there's no easy technical test to definitively prove FSO is worse than exclusive fullscreen in any/all circumstances and in what respect.
All I can say is, again, as I told @TTT, I subjectively haven't felt a difference in more recent Windows builds, and, further, it has never visibly increased input lag in any of my previous high speed tests.
Yeah, people need to realize that FSO may or may not be "worse" and depends on various factors like the game itself and other settings (e.g. Gsync on/off, vsync on/off). E.g. true independent flip mode clearly should be just as good as FSE if everything is working correctly.
I think the only way to really tell quantitatively what is better for a specific game and system configuration is to use something like GPU View or PresentMon to compare FSO on/off.
However, in my case with Apex Legends, it's just too obvious that mouse "feel" is much more responsive and I know for a fact that DWM is bypassed in FSE requiring FSO off.
For competitive PVP FPShooters I always aim for lowest latency possible, incl. swapinterval 0, ensure bypass DWM, vsync off and tearing is ok, minimize cmd and other buffers, plus many other optimizations. However, there is a trade-off between lowest latency possible settings and frame rate performance/smoothness (affecting motion clarity) which can clearly affect a person's gameplay performance too.