EDIT: false alarm (script bug; see discovery post below)
This is something that I was seeing, but not registering, while concentrating on the films’ alignment and the missing frames on one film or the other. I came across it again on a wide brightness variation across a cut. Yes, it’s a problem and it needs fixing … across the entire film!
In this exemplary sequence of frames, with different brightness levels in different parts of the picture, it is easy to see that only the frame immediately before the cut is ghosted with the incoming frame. (This should be impossible with printing equipment that made these films, but is possible with software in the scanning and post-scanning processes [stabilization, et al].) Also note that the reverse is not true – that is, on the other side of the cut, there is no ghosting from the outgoing frame onto it. And this phenomenon occurs on both 35mm and 16mm, which further reinforces the supposition of it happening from the scan on forward.
Please check this out on the actual film and tell me I’m wrong, that it’s Avisynth, or Media Player Classic HC, or undiagnosed double-vision.