If you think the Simulation Hypothesis has any merit whatsoever, you should consider the possibility that the Simulation does not have a fully-consistent engine, just one that was "close enough".
Note, of course, that the same could be true in a coincidentally-generated universe, as well as a deistic or theistic one (the latter two being effectively indistinguishable from a Simulation).
It must have a pretty good physics engine since we don’t see any strange artifacts in the sky from the times pre-cosmic inflation. In particular, it can’t be very quantized or we’d see giant “upscaling” artifacts.
It also appears to give us exactly enough computational power to build quantum computers, but not any more than that. That seems like a pretty inconvenient choice for us and the simulators.
I'm not personally all that convinced by the Simulation Hypothesis, but I recognize it's a legitimate possibility.
... right alongside the possibility that base reality doesn't have a fully consistent physics system. So far assuming reality works has panned out pretty well, but we can't be sure that will just always be true.
Not sure it's worth thinking about. You would think that a sufficiently powerful simulation could procedurally generate an arbitrary level of detail so there wouldn't really be gaps.
Which may not fit with any hypothetical Simulator's goals, anyway.
Most games don't feature an absolutely perfect Newtonian physics engine. The physics is just a means to an end, not the fundamental point of the whole endeavor.
Note, of course, that the same could be true in a coincidentally-generated universe, as well as a deistic or theistic one (the latter two being effectively indistinguishable from a Simulation).