Except for both Bell's paradox and the recent Weigner's friend variation, superdeterminism is one of the three possible ways to resolve the paradoxes, so the notion that free will exists is very much not physically clear at all given the most recent experimental results.

Also, you seem to have misunderstood my point.

I'm saying that tracking non-deterministic state changes is easier in discrete data than continuous data, so if the universe we are in is one that was designed, the design detail of interacted with quanta resolving from continuous to discrete behavior at the point of interaction strongly lends itself to the rejection of superdeterminism.

There's no advantage to switching from continuous to discrete tracking at the point of interaction if interactions are entirely deterministic, and inconsistency between the two introduces unnecessary and unexpected side effects.

The quantum eraser behavior is pretty clearly in line with a lazy optimization at work, so this conversion is apparently expensive or undesirable enough to need to be optimized away from when possible.

Modeling a continuous universe (in line with general relativity) at macro scales but switching to discrete at micro could be advantageous for both deterministic and non-deterministic simulated systems running on discrete hardware; however, switching from one to the other exclusively around measurements and interactions rather than uniform discretization across the board would be a very bizarre design decision, no?

The general difficulty in calculating certain deterministic functions which you bring up is a non sequitur to my point, unless you can make the case that converting from continuous to discrete at the point of measurement/interaction would be advantageous to an entirely deterministic system where multi-body interactions still seem to occur continuously without issue prior to measurement?