[greysphere]

I agree the model quality expression model suggests both those resource distribution models would be maximal. My supposition though is resource distribution is going to be power-law-ish, which it doesn't have to be restricted like that, but seems a natural expectation.

And yes when the available resources are below the saturation point, more analysis will be necessary to determine what's optimal. A more advanced model would account for a variety of qualities, each with different importance and different Ks, suggest distributions for those and relate them to total available resources. My intuition, based on 'lots of distributions end up being gaussian' is that equality is maximal except when significantly resource constrained (ie some state like war or famine), or k for some particularly important quality is extremely high (which interestingly is another common narrative: Noah's Ark, the Manhattan Project, Armageddon, The Martian.)