That's quite the opposite. That's exactly why I don't like Harman target.
Which is more correct?
Only way that you can do yourself is to compare directly to eq'd speakers. Then you will realize how accurately er4 represents the high frequency of the real speakers.
How?
It's an error done in Harman's research. The cause is the discrepancy between human hearing and measurement equipment. The Harman target is actually highly highly smoothed which you can't find the reason by itself. If you actually measure and compare yourself, you will notice that usually there is a boost around 8-10khz and a dip after 10k. When smoothed, the excess energy at 8-10k got spread to 6k and filled the dip after 10k. So eventually what you get is excess energy around 6khz and too much roll off after 10khz.
So what you heard is exactly why er4 has been superior and will be superior for very long period of time.