If I understand this you're saying that the DOF is so flat that they eyelash is in focus but the eye itself is not. It stands to reason then that no other part of the face was in focus. So if in fact the eye had been in focus then nothing else on the face would be in focus. I'm not a portrait photographer but I have looked at quite a few portraits through the years and I don't recall ever seeing the eyes in focus and the rest of the face being OOF. What am I missing here?
On the other hand I may have had it wrong all these years with wildlife. I thought the rule of thumb was that "if the eye is OOF nothing else matters". Now I'm wondering if I've missed the boat all these years and misunderstood what should be "if the eye is in focus then nothing else matters"?
I'm beginning to miss the film days when ignorance was truly bliss.
Caveats: I just got this lens a few days ago, have never had such a wide aperture before; nor such a sharp lens. Also, I don't do formal portraits. Therefore, I think my intuition about this lens will need recalibration as I gain experience with it.
Lens is the Nikkor Z 50mm f1.2 S. Yes. At f1.2, it is reported that you can see focus changes over about 1/4" with this lens when focused at about 26". This is by measurement, since calculation requires assuming a CoC (Circle of Confusion) which is not known a priori (at least that I can find). It has a close relative, the manual focus Nikkor Z 58mm f0.95 Noct, which has such a shallow depth of field that it is almost impossible to use wide open with a live (living, breathing, moving ever so slightly) model/subject.
My application for this lens will be environmental portraits (musicians, artists, craftsmen, academicians, mechanics...) in low light but I have only fired a few tests shots with it so far. You are right that there is a school of thought that says that as long as the eyes are in focus, it all good. I know from painful experience that some "advanced middle aged" people do not appreciate the acuity of modern Z lenses when they see the results . Who knew? So one of the reasons for acquiring this lens is, indeed, to blur facial features slightly. I've also noticed that "perceived" DoF increases with low light photography, probably because the noise masks the falling off in focus. I don't know this will actually work out for me with this lens, again because this is my first experience.
Also, despite my complaints about the Z7ii' eye detect, my belief is that its focusing accuracy will be up to the task with respect to this lens. Ya just gotta tell it where to focus. Time will tell.
* Before I give you a YouTube link that talks about this lens, let me first caution that this vlogger has gone to the "glamor side" of the spectrum (not this video, though) and some ads, or if you poke around his site, may be NSFW. *
Here is a video by Matt Granger that goes into considerable technical detail about how this lens behaves in portraiture applications. The most relevant part for this discussion starts at 5:19, where he begins shooting a sequence at f1.2 and observes the effects of increasing f-number on DoF over the face/head of a model, focused on her eye..