I think it would be infinitely more "weird" if Navier Stokes perfectly modeled fluid! These are approximations of aggregate behaviors of atoms. If atoms aren't involved in the equations, then you should expect divergence! ;)