The difference is that .sort() only uses the sign of the return value and ignores the magnitude. Ultimately, all the sort algorithm needs is the boundary condit

The difference is that .sort() only uses the sign of the return value and ignores the magnitude. Ultimately, all the sort algorithm needs is the boundary condition, e.g. a comparator. The SDF equivalent needs both the sign (is this point inside or outside of this implicit surface), alongside the magnitude, to determine the closest hit. I don't see how they're that interlinked other than saying that signed distance functions are a subset of comparators.