I made the claim As long as any satellite trails (or cosmic ray hits for that matter) do not intrude on the object or its immediate neighbourhood, our work is completely unaffected. Here is some supporting evidence which turned up quite by chance.
Below is an image from my work on U Leo and is the average stack of 32 exposures each of 30 seconds duration, two of which caught a very bright satellite in transit. The gap in the trail shows the movement during period the CCD was downloading its data between successive images. Even when diluted 31-fold the trail is still very obvious. U Leo (the central marked star just to the right of the trail) and it comparison stars are identified; none lie anywhere near the trail so photometry is unaffected. The VS is measured to be V=17.52 +/- 0.05 at 2020-01-21T00:30
