This is an interesting new take on modelling quantum correlations by analogous classical point mass distributions. It might eventually lead to a better understanding of both but for now it is just a curiosity.
formatting link
The actual paper is not behind a paywall for once:
formatting link