It’s an irregularly shaped body so there’s not one single radius. Mean radius is always going to be an approximation (even for Earth); the mean radius of the Moon is 1,737.4 km.
But it's still an approximation, not an estimate, right? (The fact that you can list the mean radius with sub-decimal precision suggests as much.)
To me, an estimate suggests that there's error bars; an approximation suggests that there's variance that we can quantify (or at least we're very confident about our error bars).
Sorry, missed a "large". It would sound pretty strange to me to claim that the distance between New York and London is "an estimated x kilometers" (with single-kilometer precision), even though there is tectonic movement etc.
