I've applied the same method I talked about in #1127675764356173885 to Jupiter's moons (except europa) and it worked quite well!
I think the processing here can be improved a lot but here are the temporary results: (I've used stellarium distance to compute the diameter from the angular diameter)
Io: 3632km measured/3643 real
Callisto: 5122km/4820km real
Ganymede: 5494km/5262km real
The PSDs with the wider circles are the smaller moons and vice versa