Hi Amirm,
I was surprised looking at your measurements. Stax is a serious manufacturer and the 006ts spec states that distortion is less than 0.02% THD @1kHz @100Vrms. So I wanted to make some measurements too.
I measured my Stax 006ts while my SR-L700 headphone was attached to it, using the second parallel output connector. So I used the real amplifier load.
My gear is not feautered for audio (Picoscope 5444D MSO using pseudodifferential setup) but it is enough to distinguish that level of distorsion.
The result is that the system shows no more than 0.025% of THD @50Vrms and no more than 0.035% of THD @100Vrms. Part of this distorsion is due to my poor setup. So I can estimate that the effective distorsion level is better.
These electrostat systems, as you know, need very low current level, so imho there is little point in measuring without considering the amplifier + headphone system as a whole.
(P.S.: the second harmonic is almost perfectly canceled. The distortion is most of the third harmonic).