Assuming the Super 8 Canon film camera as reported in the aforelinked paper all of my questions could be answered:
- Newton’s rings
- n/a (Newton’s rings have no particular expected orientation)
- Newton’s rings, being a thin film defect, are not expected to show up in multiple frames in the same way and indeed may show up in only one frame. So Stanford could just be engaging in motivated reasoning/confirmation bias – perhaps extreme.
Moreover, they would tend to show up in a fresnel-like pattern if due to a point pressure distortion in a thin film, and although it isn’t obviously a diffraction rainbow pattern in the color image, it is plausibly there.
On the other hand Faraday Effect does vary by wavelength so it may produce a rainbow pattern as well.
If it isn’t too hard, maybe I’ll try to model the Faraday Effect in Mathematica of light passing through a dipole to see if the resulting polarization results in something that could plausibly be concentric ellipses. It may be that it can’t even form that shape.