Back to the OP.
I'd start with monitoring clocks. It might even be that 2D at 240Hz is okay, 300Hz is not then you can decide whether that is worth it.
While running at 300Hz makes your GPU consume more - pressuambly mostlyl due to higer VRAM clocks - sometimes you override this.
That is, create a new 2D profile with a lower VRAM clocks. Finding a lower VRAM clock for 2D/desktop use is certainly a matter of trial-and-error as too low may lead to artifacts but often the driver team set it too high. So 2K/4K @ 60Hz might be 150Mhz or, there might be 2nd profile for - for example 2K @ 120Hz - which pushes VRAM to 300Mhz, but often after the 2nd or 3rd "desktop 2D" profile, the driver team will run VRAM at the max default.
Been a long time since I tried this - think it was in the days of Radeon HD7950 which had crazy "idle" clocks - so unsure whether the drivers can do or you need Afterburner or similar. I think the current AMD drivers might be able, unsure about the Nvidia ones. Might be too much effort to save 10W or so but making things quieter is always nice IMO.
Not many places test this kind of thing. I though computerbase did but I could not find it their 5070 Ti review so maybe I mis-remembered or they no longer do. But even then, I cannot see any place doing much more than the basic - TPU only do single and multi-monitor.
Take that back I did find it in the CB review:
Nvidia GeForce RTX 5070 Ti im Test: Lautstärke, Temperatur, Leistungsaufnahme und OC / Lautstärke & Kühlung
www.computerbase.de
So their dual monitor does show 2 x 4K @ 60Hz and 2 x 4K at 60Hz + 144Hz. The 5070 Ti does pretty well there "only" increasing by 10W for the later.