I do think the built in margin over stall for the published Vx is most likely. Perhaps in Aviat's opinion, Vy doesn't get close enough to stall speed until you're high enough to be well clear of terrain in most cases to bother applying the same correction.
I really don't think that makes sense. It is not uncommon for Husky owners to operate into bush strips at density altitudes of 10,000 plus and the company is based in the Rockies so I doubt they would assume 10,000' is going to be well clear of terrain. Anyway, the clean stall speed is 58 mph whilst the SL Vx is quoted as 67 rising to 70 at 10,000' ie the margin gets greater with altitude, which is the opposite of the above rationale.
If it was about choosing a Vx to increase the stall margin - which is hypothetically plausible - it would defeat the object to retain a Vy for 10,000' which was slower than a Vx that had been chosen to be a safe minimum. Quite apart from the absurdity of recommending a Vx which actually results in a lesser climb gradient than the corresponding Vy, particulalry in an aircraft we know is often operated in mountainous terrain at those altitudes. Considering that there are at least 3 anomalies existing in this AFM around Vx, it is obvious that they have simply made a mistake. These things happen.