Unfortunately we found that most recordings have a treble boost of about 5dB at 10kHz to make up for absorbtion in most rooms when listening to loudspeakers. We have data on our web site from five of the studies that we found in the literature. We did a study for an AES meeting at a recording studio in Chicago to determine if our estimation was correct. The results compiled from the 20 people in attendance verified that 5dB of high frequency cut was appropriate.
I was waiting until I had more time to talk with MR. Linquitz about his experiments because it will take some time to explain and dig up the backing data. What Mr. Linquitz did is make himself a personal HRTF that sounded correct for himself. The equalization that he thought sounded correct might be quite unappropriate for others. The difference he finds at 2.9kHz is most likely a slightly different ear canal resonance. I am not sure what the need for the sharp dip at 7.8kHz is about. If it was in the 10-12kHz area where we have a bit too much energy I could easily understand the cause.