The problem is that currently particles/dust (in rings, galaxies...) are masless points, so they can't collide like normal bodies do when they merge, and they don't have any mass. If they had mass, they would slow the simulation about as much as normal bodies, I think, and then they would be pointless (now in both senses of the word).
Particles would probably need to have a volume to make it possible for them to combine, or to determine if they should combine to some "bigger" particle or a massive body depending on how close several particles are to each other.
I don't know how they could determine the mass either, when the particles could represent anything from dust in planetary rings to stars in galaxies. Maybe based on the body they orbit (if they orbit any)?
There could be an option to choose how much "potential mass" a ring could have, and then each particle could have that "mass" divided by the number of particles. But the mass won't become real before it becomes an object, which it does after several particles have merged... And all the "potential masses" of those objects have been added to each other as well. Then the program just has to decide what a collission is when it is between points with no volume.