In the business of space visualization, a well known problem is that of precision at large scale differences.
The short answer is: no, this asset is not designed to allow arbitrary scaling and floating point precision.
Naturally, if you let a unit in unit be large enough, preserving precision for small scales is going to be a problem. The same will apply for physics calculations in a single scale of reference, but far from the origin.
The ray marcher doesn't care much about scale in the first place, and precision inaccuracies would normally not cause any visible artifacts in the galactic cloud rendering. What you may see is precision issues due to the limitations of the input texture of the galaxy, if you up scale enough. but you most likely won't notice this either as observing any issues probably will require the camera to go inside the cloud and at that point you don't have a clear frame of reference.
The star field renderer might run into some issues with precision if you zoom in on their positions enough in a galaxy that has an immense scale however. Perhaps zooming in on a generated star system while still working on the galactic scale could cause issues with your planets or asteroid positions, it's hard to tell. It's really up to your use case.
There are several ways to solve this problem, and we might decide to employ one of them somewhere down the line.
However, all in all, our recommendation is that if you're working on a game, you're much better off faking it. Re-scale and re-center, perhaps create a transition effect between scales or employ multiple cameras which you can fade between. The cases where your users are going to care how you solved this issue are likely extremely few.