Sure, I am aware of that. But in this forum, I have seen many instances (like in this thread) of it being used to do all sorts of things: compute the dielectric constant of graphene, salty solutions with an external electric field, iron oxide, the diffusion coefficient of CO2 in CaO (okay, for this one a reaction may occur), just to quote the most recent ones.
If one is interested in modelling some physical observable depending on the microscopic distribution of particles, then there are much better choices.