hmm a good 4dB level difference between L and R from 300Hz to 2kHz (Tyll's version).
20Hz = -2dB drop off above 10kHz
And to think that some people are bothered by a 0.1dB channel imbalance from an amplifier

.
the GE measurement seems to differ a lot (I know...) no channel imbalance there.
http://en.goldenears.net/1068020Hz= 0dB drop off above 10kHz
The Russian guys also had a better looking FR
http://personalaudio.ru/raa/otchety/naushniki/beyerdynamic-dt-250-250-ohm/BUT quite some differences between L and R also in the range of 4 to 5 dB which is (should be) more than obvious.
20Hz = -10dB no drop-off above 10kHz (a peak is even visible)
The Headroom plot looks a bit suspect .... roll-off in the lows starting from 100Hz and in the highs from 10kHz onwards ?
What's with that ? 20Hz = -15dB drop-off above 10kHz
I think the title of this thread thread is right on the money.... inconsistencies all over the place (channel imbalance) and or measurement/compensation errors.