Great rabbit hole, very interesting. I'm not quite wrapping my head around this from a quick glance, though. So you take a series of images at different focal depths, and this software can focus-stack those images to take the sharpest features of each and create a composite. I see it also creates a depth map of the image, which is super cool and has lots of applications. Is the stereogram generated directly from the depth map and value data? So the image is taken from a single point of view! Then inpainting would be necessary to synthesize offset views? But the image looks really good, perhaps it's just hard to see the artifacts?

At least now I understand why my macro photos are so much shallower than some that I see no matter what I do in camera.