In addition to Mark's analog video image, I found images of the M87 jet by AP folks bobzeq25 and schmeah in this related CN thread, which used small scopes and low magnification with image scale courser than 2 arcsecs/px (bobzeq25 used a 70mm scope at a Bortle 7 location, and schmeah's image scale was 3.4 arcsecs/px, see "example" at end of this post). I also measured the distance between the nucleus of the M87 galaxy itself and the brightest spot of the jet closest to the nucleus to be at least 10 arcsecs (using this, this, this, and star charts).
So it now seems to me that it was probably wrong to think along the lines "Since the jet is very small (~1 arcsec?), we need a sufficiently fine image scale to see it". Perhaps a better way to look at this might be the following:
[Aperture and image scale here may be red herrings:]
The M87 galaxy itself is a relatively bright source of light with an intensity distribution spread spatially. The jet (or rather its bright point close to the galaxy) is also another light source, with another spatial intensity distribution. The distance between the peaks of these two light distribitions is at least 10 arcsecs, many times greater than the diffraction limit of even the smallest telescopes, so aperture should be less of an issue in "splitting the two peaks". Also, the image scale need not be as fine as I had initially thought; perhaps it is enough to be less than, say, half (Nyquist?) or one-third of the distance between the peaks. Of course conditions/seeing will be a big factor, but suppose we have good conditions.
[The key factor may be the amount of dynamic range available within a certain luminosity bracket:]
A main difficulty in seeing the jet seems to be that the spread of M87's stronger-intensity spatial distribution tends to swamp that of the jet. (This is just qualitative; to state this precisely would need quantitative/FWHM measurement data over the visible wavelengths.) So to clearly distinguish the peaks of these two distributions it may be useful to capture using an appropriate "intensity bracket". (EAA/AP folks will know better about how to do this; may be one needs to start with an optimal low gain setting and stop sub-frame exposure at an optimal point of time; sounds difficult.) Also, this means that the dynamic range of the sensor can be an important factor (dynamic range = well-depth/read-noise; the ADC bit-depth can be a rough indicator of dynamic range).
An example AP image where the jet is caught at 3.4 arcsecs/px:
This Markarian's Chain image (AP, processed, not EAA) by Derek Santiago (CN member schmeah) was taken with a small 85mm f/5.3 scope, and a 7.4um pixel size camera, giving a "course" image scale of 3.4 arcsecs/px. It is a 2048x2048 image over a 2 deg x 2 deg region of sky, where M87 lies in a small crop area. Yet, it captures the jet remarkably well! (I needed to adjust my image viewer's curve for luminosity-value/gamma, and then the jet stood out stunningly, pointing toward 5 o'clock.) Note that very high quality equipment were used: A Takahashi FSQ85 scope on an Avalon M-Uno mount, and a QSI 640 camera, which has CCD sensor and 16 bit ADC.
Edited by adg, 19 May 2019 - 05:07 PM.