> It's worth pointing out that particle-based fluid simulators are not physically accurate.

Do you have some references for this? I was under the impression that convergence results have been shown for purely Lagrangian approaches like SPH.