It would seem necessary to know the projectile’s velocity at known distances. Labradar or multiple chronographs along the bullet’s path…
True, and air density conditions have to be accounted for.
BC is tied to form factor w/resp to drag tables, or a drag rate, as determined at a standard atmosphere, and at each velocity point.
For example; Berger uses ICAO, Sierra uses Std.Metro conditions.
But because their bullet drag curves(actual) do not match gov't standard curves, BC changes at different velocity points. Sierra identifies several velocity point differences, Berger chooses a reasonable average.
If the curves matched, maybe you made your own custom curve, then BC would hold as a fixed term.
Most software I've used converts the standard BC you've entered to an internal 'local' BC with air density changes.
So to test and declare BC takes some careful accounting.
For most of us, it's just shooting a bunch of ranges and adjusting BC for best match.
But local air density and MV at least, must be accounted for.
Seems like some bullet makers just go out back and shoot different ranges, and adjust BC in their software to reverse engineer a BC value(like we might), but failing to account for local air density conditions. This would lead to bad solutions every time.