This seems like a good approach. Often we either have a different set of underlying assumptions/facts or different axioms. For the former case, the “how truthful” approach makes a lot of sense.