This is a follow up question to What exactly is a "moon"? The conclusions I draw from James K's answer is that the IAU should define what a moon is. They haven't done so yet, but they should.
The obvious way to go about defining "moon" would be to build on the definition of a planet:
- Planets must be in orbit around the Sun
- Planets must be round due to their own gravity
- Planets must have "cleared the neighbourhood"
This immediately suggests the equivalent definition of a moon:
- Moons must be in orbit around the host planet
- Moons must have [minimum size]. Unfortunately being round doesn't work since not all moons are round (c.f. Mars' moons, unless one reclassifies those as not moons)
- Moons must have "cleared the neighbourhood" of their own orbit, i.e. in their particular orbit around the host they are by far the most massive body
Would such a definition work? If so, why hasn't the IAU also defined moons? It seems so natural after all. If not, what's the catch?