Hej,
nice starting point, maru.
I would also recommend using AO with inverted normal "mode" for saturation (desaturation).
As you can notice on romullus reference, this is important.
Also the very bright reflection parts in the reference are very important.
But I now have no idea reproducing this effect.
Maybe another approach would be some kind of vector displacement.
As the material exists out of small blisters, that are truncated..
Hope it helps :)
All the best