I wrote life simulation and some very wimpy video games way back on my TI99. A game such as asteroids has limits. Boundaries. The machines do not have infinite memory or register size. You use arbitrary code limits to wrap around a finite space. So it is not a good analog for infinity. They are loops.
Everyone disagrees with that idea. The surface of a sphere is finite. And yet, despite being finite, it has no boundaries.
I can agree to a certain extent that it has no boundaries. But not totally. I guess it is what the definitions of boundary are. Can you define the circumference of a sphere if you indeed have no boundaries? Or is there the boundary of the start and stop of a distance round the circumference?
Linguistics bumping into other things.
Sure. Nobody says points on the sphere are indistinguishable. Rather the opposite. There is a length metric available to you, and lots of other useful things. There is a notion of straight line, and of angles. You can determine the circumference by working to find the longest straight line that returns to its origin. The properties of geometry will be a little different to what you might be used to, but they make sense.
In all, most people are pretty happy working on the surface of a sphere navigating themselves around their daily life. Your car GPS and maps for one example. Height variation of terrain can be seen as just a weird local warping of distance.