Why does generation not matter? ps4 and xbone came out at the same technological timeframe. You should't let your brand preference limit you to using the xbone as the basic metric.
And what world are you living in? There are multiplatform games, running on both ps4 and xbone. Those that run exclusively at 1080p on ps4, will not run at 4k on the xox at the exact same settings.
The point is that across different platforms there are no "exact same settings". You have two unique implementations of the same game, but they are not in anyway compatible. This means that games can run better or worse on each platform than the paper specs would suggest, and because you can never run the other version on the opposing console you could never benchmark how it would actually perform. Considering this is a Quantum Break thread, let's use it as an example. This game (and many others on PC) has both a DX12 and DX11 implementation, and they perform nothing alike... on the exact same specification PC. You can't do "like for like" comparisons across incompatible platforms, because the software stopped being like for like in order to even run on each.
Honestly, I'm surprised you're struggling with this. Your whole argument was about having a defined actually testable metric for a term... and you've since only been arguing for the untestable.