If we're going with the theory that things with a wavelength shorter than the Planck length don't exist then the sampling theorem should allow us to describe the universe with a discrete set of points.

General relativity complicates things (as usual) but the sampling theorem can be generalised to non-uniform samples so it should probably still work somehow.

In both cases this discrete theory is equivalent to a theory based on real numbers though, so in that sense there's not really any way to say which is better.