67P/Churyumov–Gerasimenko has a highly irregular dumbbell shape. But the sample of comet shapes observed is very small, so I wonder if irregular shape is the norm for comets and for small moons. Many known moons are no larger than this ~4 km diameter comet. The smallest moons imaged are modestly irregular, basically just elongated, at least they don't have a waist like 67/P. The most elongated moon is perhaps ~135/60 km Prometheus.
Is there reason to believe that moons which are too small for hydrostatic equilibrium get more rounded than comets of similar mass and composition? Moons differ from comets in several ways. AFAIK: Moons in general have a very different gravitational environment, a much more stable distance to the Sun, experience more frequent impacts, another formation history if not captured. Composition and density depends on the formation distance from the Sun, although even ~300 km Hyperion has a similar density as 67/P.
Should we expect the small moons in general, and the moons of Pluto especially, to be rounded or dumbbell shaped? If 67P is a merged binary, isn't it more likely for two objects to merge if they are moons instead of comets, since neighboring moons have low relative speed? ~100-200 km Janus and Epimetheus look as if they could merge fairly calmly. Does the lack of (observed) dumbbell shaped moons tell us something about 67P, that its shape is a rare outlier for example?