Here is another test shot with the ASI462MC. In this case there are no filters (except the UV cut glass in the camera body).
The light source is an 850 nm LED. The observatory is otherwise dark. Blue values (at even,even coordinates) run about twice G and R values.
I don't know how to interpret this other than to say that either I'm doing something dreadfully wrong or Sony has misrepresented R and G transparency at 850 nm.
I understand NIR astrophotography is a narrow niche, but I hope no one thinks the IMX462LQR color sensor is right for the job.