Float Space Precision Improvements

Are there any potential ways of improving Blender's float space precision issues? It would be really nice to work in bigger scenes but it's super difficult with most DCCs. Is there any known workflow worth implementing into Blender? I know Houdini has something called double precision and I know that game engines like UE5 can shift the origin between world tiles.

In an extreme case, they are also custom game engines that use 64 bit precision for super large scenes. Is any of this possible to work towards to for Blender?