Some comments on the paper “Est Modus in rebus: analytical properties of multi-model ensembles