Right, of course you aren't going to get good results if you aren't identifying and targeting distinct problems, you need to know what your target is before engaging.
I'll try using my EQ tuning of the QDC V14 as an example to explain my strategy.
I used 6 filters here to address certain problems.
For the bass region (300Hz and below), I'm using two low shelf filters to shape the bass according to my preference for sub-bass (lots of it as you can see). On a HP this bass region would be completely nonsensical because the amount of kinetic energy generated by the driver is much higher in general, so less measured bass is required to create "correct" perceived bass levels. IEMs generate much less energy, so listeners consistently go for higher levels of bass (as seen in the Harman IE target) in general to compensate for the lost sensation of bass going from HP to IEM. Whatever your preference for bass is on HPs, add on significant amounts of sub-bass. I reduced the amount of mid-bass here because that stock amount was masking the pinna gain region (1k to 4k) too much for my liking, so I nudged it down 1.5dB to accentuate the pinna gain region's prominence in the tuning. Generally, you will want the overall level of mid-bass to be roughly within a 3:5 ratio to the pinna gain peak to avoid masking while still having sufficient body.
Mids before 1k generally form the baseline for the tuning as it contains the fundamental frequencies of much of modern music, so generally a flat response from 300 to 1k is preferable depending on how early your pinna gain region kicks in.
My pinna gain region (1.1k to 3k) is a problem area on most IEMs, so this area needed 3 filters to shape to my preference. First is a slight nudge down at 1.6k because vox stems and guitar harmonics were coming on a bit thick in the mix, this usually denotes an unbalanced ratio between the lower and upper pinna gain region. This of course then points the finger at the elephant in the room that is the notable notch at 3k. This notch causes me to perceive an elevation drop on the center of the sound image, which I fixed with a narrow 3dB peak hitting just that notch. I arrived at that level by ear, paying attention to the center image until it was at eye level. The last filter I used here was at 4.4k to address some mic buzzing and digital quality that was coming through on tracks. This region (4k to 5k) generally needs to gently slope down from the gain peak to help control odd harmonics that are generated from the 1k to 1.7k that cause a metallic tinge to permeate your mix, so you want to target that region whenever you hear that kind of noise.
The last filter I applied is at 8k to control some excessive sheen and borderline sibilance. This region is a common resonance point in measurement couplers so take the measured level with a grain of salt, but the resonance points of the human ear typically occupy the same area, so you might have to hunt around a bit to find and fix sibilance points on your tunings.
