The face exists on two axis! The top half of the panel the head is looking a little up and off, the bottom half of the panel, the face is looking down and away. The act of rending out the image in a more realistic painterly way made this error very clear. You looked at it and it screamed, "nope".
So why does everyone like the panel so much if, in fact, it's drawn completely wrong? Is it because that is line art in comics is more forgiving than something that looks more real? Well, that's part of it, for sure. But it made me think of the King of drawing out-of-proportion, Jack Kirby. Kirby's aesthetic of movement within panels is legendary. People literally jump out of the page, Jack Kirby's dynamic action set the template for how Marvel comics should look well into the 1990's. A lot of his work has a goofy sense of perspective, but it still works. And his seemingly all-done-wrong art, which had more dynamism that anything being done today. Is doing it wrong the key?
When looking at my own panel and the two axis of the face it made me realize that maybe the static-in-time images in comics are not as static as we think they are. Perhaps my initial panel worked because it was wrong. That the subtle difference in axis of the face created not just one face, but a composite of two images of the same face, spread out over time. So when you read the panel, it is not just a static image in time but subtly animated. The subtle wrongness of the image gives it more movement, your brain fills in the gaps, so to speak, and you see her lifting her head slightly over time, even though the image itself does not move. The contrast of the image, white above, black below, might also be helping with this faux movement, drawing the eye along in the direction that the head would be moving in, so by the time the eye gets to the end of time represented in the image, it has actually moved forward a second or two in time, and the previous part of the image left behind is almost ignored by the brain that it is not quite perfectly in line with the "start" of the image.
If this is what is happening, Kirby's work was at the extreme end of this effect. Where my panel may be a second or two, Kirby's panels existed sometimes with a six second gap between foreground and background.
It's an interesting idea that I think merits some experimentation.