Yes, it IS a referee problem. It was a huge impediment to letting players loose with a ship in 2300.... at least for me and a few friends who ran it.
It was also a referee problem in FGU's Space Opera.
And in ICE's Spacemaster.
I don't have any others coming to mind with full 3d mapping, but I'm certain I have at least a couple which are not memorable and have xyz coordinate systems in use.
It was a huge issue in 1987, when most calculators had 5 math functions ([+][-][×][÷][%] ) and 3 non-functions ([C][CE][=])... only a few had a [√] without being full fledged (and expensive) scientifics...
I did have one, but didn't think to use it. I just ran my 2300 games either in the OTU (rather than the 2300AD setting), or ran th e2300 setting on a single world per campaign.
Later, I used a square root table, once I finally grasped the hypotenuse formula doesn't care how many dimensions...