So frame generation and the input lag it adds is literal Satan, out of the question. I'm only interested in upscaling solution for games that don't have good support for native motion based upscaling like FSR, Xess or DLSS.

However as I understand this generates a second window. on top of your game.

From some further reading it seems the actual game window isn't running in an optimized borderless mode because It's not on top and G-Sync is turned on for the Lossess Scaling window rather than game window.

This seems like a such a cursed way of doing things if you care about latency at all. It makes me wonder why would I want to use this instead of a simpler post processing effect that injects directly into a game instead.

Or am I misunderstanding what this software does and it's use case in general?