well, to
@GaryH's point, individual loudness contours and HRTFs vary like crazy. The harman curve is meant to approximately compensate for the difference between a headphone's sound and an idealized speaker's sound in-room. But it's not possible that a standard curve will fully achieve this goal for every ear. It can compensate for the physical and audiological / psychoacoustic things we all share in common. But unfortunately those commonalities don't get everyone the sound they want.
In fact, I spoke to an engineer at a very well-known IEM brand (you've heard of them, maybe even your mom's heard of them) and they told me they had tried to categorize the pinnae of real people. The idea was to measure enough of them to come up with groupings / types of ear shape so that they could in theory customize sound according to your "ear type", something along those lines.
They gave up on this effort after finding no pattern in ear shapes after dozens of measurements.
At the end of the day, the harman curve is a good starting point, but adjusting EQ by ear (NOT via tones, obviously, that's only for arrogant fools) seems to be the only option for a final finishing step between "good" and "perfect", if any such thing can be said to exist in headphones.