Really interesting question. But post #2 says it can happen on a 2 blade RC model and a quick search on internet even gives a video demo of it on a Raptor 30:
The natural frequency on the RC model looks to be quite high compared to my memory of seeing the onset of ground resonance on a full scale Squirrel many years back.
There must be mass and stiffness scaling effects that come into play when considering RC models and their full scale counterparts. For roll motion of a helicopter on the ground, natural frequency is determined by the square root of (roll stiffness / roll moment of inertia). I suspect skids of a model helicopter are
relatively much stiffer than their full scale counterparts. The roll moment of inertia probably remains reasonably in proportion with the size of the model or helicopter. So that would be consistent with model natural (or resonant) frequency being higher than on full scale.
The blade grips on a model helicopter not only allow lead and lag movement of the rotors, but I guess also provide some friction damping. Didn't really old generation helicopters also have friction dampers for lead and lag blade movement?
Aside from the model blades getting stuck in their grips so they are not exactly separated by 180 degrees, there could also be imbalances in the stabiliser bar if not set up properly. If anything, I might have expected ground resonance to be more common on models than (airworthy) full size helicopters.