How large is "large $N_c$" for Nuclear matter?

We argue that a so far neglected dimensionless scale, the number of neighbors in a closely packed system, is relevant for the convergence of the large $N_c$ expansion at high chemical potential. It is only when the number of colors is large w.r.t. this new scale ($\sim \text{order}10$) that a convergent large $N_c$ limit is reached. This provides an explanation as to why the large $N_c$ expansion, qualitatively successful in in vacuum QCD, fails to describe high baryo-chemical potential systems, such as nuclear matter. It also means that phenomenological claims about high density matter based on large $N_c$ extrapolations should be treated with caution.