Dude, that was a trick question. If the scale came first then that's a classic "fitting the data" science FAIL
Nah, it's actually a colossal fail because that just enables all kinds of bias given all that other shit dude was doing especially since the the scale would have to be arbitrary as fuck given the ambiguity of the phenomena he was testing.
You sure the scale came first?
I can't recall, but I think you are right. He had a fucked up frame of inquiry and completely non-scientific methodology but his theory still has merit imo.