Also the article is most likely wrong about calling these artifacts purely diffraction-related, creating diffraction images that big and visible would require too small focal ratio, which is unlikely for night conditions, because they increase exposure time which is already too long. These are just a consequence of internal back-reflections inside the lens optics which manifest itself on extreme contrast parts of image (bright light vs dark sky). Especially obvious on astronomical images, where control of focal ratio is manual and it's almost always open fully to collect as much light as possible - of course, focus depth is of no meaning for astronomy - and the quality of this lens allows for perfect images even when wide open. Bright stars always looks like this on Canon 14-40L.
i don't understand what you are saying here. astronomical images don't look exactly like that - they have fewer "spikes" - but they are similar. the spikes are either from diffraction caused by the secondary mirror support (which light passes through in a reflecting telescope) or readout artifacts in the CCD, particularly when it saturates (although that looks quite different - a single vertical line, often).
i agree that the spikes (in astronomy images) don't come from blade edges in a diaphragm aperture, but they are still diffraction spikes (from the secondary support) and "everyone" (as far as i know / remember) refers to them as such.
so what are you referring to with "internal back reflections" and how does that give spikes (rather than ghosts)?
[above from (past) personal experience as an optical astronomer on research telescopes - i am not sure if you are confused, or if there's some kind of problem i never understood. perhaps you work on a system where this is an issue?]
I am speaking about astronomical images not in general, but those taken with the Canon EF 17-40 f/4 L USM lens. They are almost always taken with the diaphragm wide open, and contain same kind of spikes -> spikes are not from diffraction.