It’s a long shot but imagine if you will a gluon made of two halves. The halves can each be drawn from the two “weights” (low and high of a non-zero-sum field) if they broke symmetry and the two “charges” like from the zero-sum field.
Given cancellation and combination, 8 gluons happen.
L+L+, L-L-, L+H-, L-H+, L+H+, L-H-, H+H+, H-H-.
So there’s more green weight “sticky” and the Boltzmann distribution for the half Bose-Einstein as a Fermi-Dirac perhaps. The blue colour perhaps travels less far due to higher “mass” (if it splits), but as the energy input in the strong force makes more gluons at a critical threshold, the further interaction has more energy and a less gluey implementation in blue.
I wonder if the QCD simulation evaluations can take this all into account for better accuracy. I put the two yellow “charges” in there which technically would be massed green, but given the charge +/- cancellation without perhaps LH equality would suggest a kind of neutral weight dipole.
EDIT: As the energy increases moving into “small x QCD” the expansion of colour coverage, to prevent saturation by a UV gluon density catastrophe, the critical temperature is exceeded and the “cooper pair” effect on the half bosons is removed and Pauli colour saturation removes the gluon density within the nucleon. Yes, this paragraph is unproven, but there must be some effect stabilising the UV catastrophe. This would also lead to a cyclic order of colour based on mass expression above the critical temperature for some critically small x.