It's not given much explanation in the paper... here's my understanding.
The paper shows that the total number of rings grows faster than the number of local rings with m²=0, but grows at the ~same rate as the number of local rings with m³=0. So, maybe the local rings with m³=0 are ~everything!