It turns out my Dandelion chip had not been destroyed after all. The plastic casing was damaged, but the chip was still operable. So I fortified the casing with some epoxy and reinstalled the chip on the Adaptall-2 mount after removing more material from the light baffle to get the chip into the right position.

I now present the results of the exposure test. I took two sets of exposures from f/2.8 through f/22 at whole-stop intervals. Manual exposure was used in order to eliminate metering variation as a factor. ISO was a constant 100. For each stop change in aperture, I simply changed the shutter speed by one stop to get the same theoretical EV every time.

The first set of seven exposures was done using a command dial on the camera to control aperture. The second set (the control set) was done using the lens's aperture ring to control aperture. Both sets can be found at the following URL:

https://drive.google.com/drive/folders/1EffrCxTsWboDwtgkyY5gcl2sevk7b896?usp=sharing

Spoiler alert: it is obvious that this lens, or the combination of this lens and this Adaptall-2 mount, does not provide linear control of aperture via the stop-down lever. (Again, I am using "linear" here in the sense that a constant increment in the position of the stop-down lever will produce a constant increment in the f-number. For example, one millimeter of change in position of the stop-down lever will result in one stop of aperture change.) You can see how the photos become progressively overexposed as aperture is changed from f/2.8 to f/16. There is a slight drop in exposure going from f/16 to f/22 for unknown reasons, but the overall conclusion of non-linearity is inescapable.

Curiously, you can also see some inconstancy of exposure in the control set. Exposures get slightly darker as you go from f/2.8 to f/22. But not enough to really matter.

I am going to keep the Dandelion on this lens but use the aperture ring since it obviously produces much more consistent results. Oh well! At least I get to retain some of the other benefits of the chip: correct lens identification in EXIF and the enabling of highlight-weighted metering.

Thank you guys for the appreciation.

  John, I like the touch of borrowed hair Your point of view made the connection!
  Fons, you have the ability to paint with blobs of color. You do it often. I like the anchor of the red flower and the wandering around with a safe place to come back.

Ok....then there is only the "paper scale" solution provided by Nikon...…
The infinity focus at 20mm is quite far away from "visible light" infinity…..so not so well corrected I guess.

Thanks Armando.

There is a bit more information about 20-35 here:
https://www.mir.com.my/rb/photography/companies/nikon/nikkoresources/AFNikkor/AFNikkor2035mmf28D/index.htm

…..and development of wide angle zooms in general. Also link to Nikkor 25-50.