The 10 dB dip at 2 KHz…. must be noticeable.
Did you actually read the article...
"While the mids are quite neutral it does (very slightly) lack some ultimate ‘clarity/presence’. The dip around 2kHz is responsible for this. That dip is smaller in reality than it measures because the so-called Concha-gain will make this measured dip smaller than it looks. It is still there though, just not audibly as deep as the plots suggest.
A dip in this part of the frequency range is more desirable than a peak."