Can be quite fraught trying to use a coupler and different tips and trying to use that as an indicator.
For example. here's my frequency plot along with some others - I think by now most people know the calibration exercise I've gone through to get close to IEC711 standard.
FH5 is in red (foam tips) and does show some differences through upper mid-range to yours Theo. The gap from 4-6 kHz is still there - but not to same extent. In graph above I also included Andromeda, 64 Audio U10, and FLC8S. You could debate that those 3 earphones are stellar performers at their price points and considered pretty balanced. Two things stick out .....
- The upper mid-range (presence area) very big colouration on the FH5 (makes it very forward - especially in the vocal area)
- All 4 earphones have a natural dip at 5 kHz
The point I'm making is that this dip doesn't tend to be noticed - if it was earlier or later in the curve, it would definitely come into play.
And then we can talk about tip differences. Definitely audible - but we can't make sweeping observations on how they will change the sound - when we are using:
- A largely uncalibrated coupler - which is very differently shaped to our ear-drums (no bends)
- Measurements that cannot take into account individual physiology
- Measurements that cannot take into account individual levels of seal.
I don't tend to use spiral dots because I personally get an inconsistent seal - for others they could be perfect. Seal will have a far greater effect on frequency response than the slight variations materials can make. So does insertion depth (and angle) for that matter.
I would argue that the point you've indicated the mid-range drop is, doesn't really matter (~ 5 kHz), the magnitude of the measured delta in freq intensity change pales into insignificance given the magnitude of the relative change anyway, and the size of the upper-mid bump will mask any tip change anyway.
IMO - use the tips which give you best and most consistent seal. If thats a Spiral Dot - go for it.