.... I think Audeze's response was a good one in this case, because people don't understand that measurements aren't an arbiter of truth, but only a guide with many caveats.
Exactly! Measuring higher frequencies (above 3400Hz) becomes very problematic to do reproducibly as even the slightest repositioning on the dummy head has huge ramifications on the measurements in this FR region (even when averaging out 5 separate positions/locations on the dummy head). I did some work with Tyll a few years back and categorized just how much this variation is...and it can be quite significant. So while the lower octaves are quite repeatable and reproducible, not so much the higher you go. I use measurements only to parlay part of the picture...and I still have to rely on my ears for the rest.
In case anyone is interested:
http://www.innerfidelity.com/content/expert-tests-innerfidelitys-headphone-measurement-repeatability-and-reproducibility#aPgTxEUSpoAh16c9.97