It would have been better if they had said, "filter correction" of 1-1/2 stops. Only 2/3 of a stop is unrealistic.
Meanwhile, neutral density is factored in logarithm mode. Every increment of .30 equates to one more stop (or EV). So that obviously can't correspond to that typo or whatever of 1.2 in the case of an X0 filter.
I've tested the X0 not only very carefully outdoors using a 1 degree spotmeter, gray card, and final bracketing comparison versus an unfiltered shot measuring the density of the developed film - I've also very critically used an expensive easel densitometer in the lab designed to be color neutral to measure projected 5000K light through the X0. It accurately reads within .01 density! The reading is taken at a specific dead-on focused angle, free of the sort of flare and cosine error one might get using casual metering. With TMax100 film, it's very close to a 1 EV shift (.30 density), with FP4, reasonably 1-1/3 EV, with Delta 100, 1-1/2 EV. I have several of these HMC X0 filters, all well cared for and unfaded, and the variation between them is negligible - only a few CC points.
Actual transmission might vary a few percent between the single-coated versus the multi-coated HMC version - again negligible when the lens is shaded.
But I still don't understand what all the confusion is about. For many years they've stated on lists accompanying the filter, or even on the plastic container, that the exposure offset for X0 is within that same very range. Why make a wild goose chase out of it?