If our framework is limited to clear distinction of the researched subjects, then:
Given some cardinal (notated as C), Entropy (notated as E) is in inverse proportion with Distinction degree (notated as D) where 0 =< D =< 1.
E = C/D and this fromula is the limited case of clear distinction of the reseached subject (where clear distinction is used as the general paradigm of the mathematical science, for the past 2500 years).
Well, at least we now have a formula. Are you willing to stand by this?
My initial observation is that this means that [4,2,2] has exactly twice the entropy of [2,1,1], which violates your statement earlier that size of difference does not matter.