That's a fascinating one. But how would one proof against a false positive of fakery w/ actual inability to hear either?
What about multiple sets with different words but still at the same frequency, as redundancy so the odds of a false positive are reduced with each additional set?