>>42664371
>whoever did that port would have to know exactly what they're doing
As jaded as this sounds, I couldn't see that happening in the timeline we're currently in. Devs on UE5, amateur or (allegedly) professional, have had an overwhelmingly bad track record of actually understanding the limitations of the tools they've been given.
While a theoretical "This Could Be Second Life In 2021 [REAL]" might be possible, I'd suspect the best we'd get is a ~10-20 FPS increase with about 8x the hardware requirement SL currently demands. How much of that is down to UE5 being itself, versus the visibly growing deficit of optimization-related knowledge shown by current-day devs, is up for debate.
Would pave over those directly imported, blatantly unoptimized (and probably store bought) models having like 300,000 polys by default though, yeah.