I don't believe the axiom of choice has ever played a role in bioinformatics. I certainly never expected it would! But it shows that monotonicity is technically needed, although only to remove pathological examples which additivity alone doesn't clear out. 23/