Interesting topic about this.
Every raw is a container with processed sensor readout data.
Some brands only adds correction data of the camera + lens combination and focallength and aperture characteristics to the exif file which a rawdeveloper can read and use for it’s own calculations to correct.
Panasonic chooses to embed some corrections, applied on the sensordata before creating a rawfile.
Fairly sure shadowcompensation, vignettingcorrection, and maybe diffraction compensation. The Airydisks thing.
Idyn, intelligent dynamic controle, is for jpeg contrast tonality ánd exposure effective.
For rawfile only -1/3 -2/3 and -1 stop exposure correction as influence.
Ires, using silkypix that reads Ires settings stored in the exif and applies that it self on the rawfile same as camerastyle like vivid or custom.
DxO can’t read those settings and does nothing with the exif info.(or won’t delibralatly)
I conclude this after some reading, every brand does it’s own rules and actions on exif and rawfiledata and every developer software does it’s own actions on those data.
In your case nikon and capture nx2 are as silkypix and panasonic/fuij, buddy’s.
So the info distrubution in the rawfile is more detailed.
Only way to get facts and clear view on what and when and how the effect is of dxo on which rawfile is testing , shoot with and without settings active in the same scene.
Then and only then you can determine which settings are idle in raw and which are effective in exif reading or even imagedata.(non reversable like shadow compensation)