This is only true if you can be sure of your starting axioms, which is not the case in social organization. Experiments simply add specific metrics to your set of axioms.

It all comes down to heuristics, and differing heuristics are not mutually exclusive.