Thanks for resurrecting the old discussion. In retrospect I should have pushed back against hank's blithe certainty that there can be 'No "slow, grazing collisions"'.

Of course there can be. Orbital mechanics can be very counterintuitive--"go faster to go slower (higher)". The example of the correct dynamic I had in mind is like the slingshot maneuver, in which one object approaches another from behind in its orbit, traveling at a higher velocity. As the overtaking object passes the other, the other's gravity pulls the overtaker; at the correct angle for a slingshot it goes careening off in another direction, but at other angles it can slow down as the first object pulls it "back". I'm pretty sure that most Mars orbital insertions have leveraged the effect to save energy.

Those kinds of maneuvers may seem improbable only because when we do them we have a goal in mind and it's hard work to accomplish that goal. But stuff happens in Nature, and somewhere sometime some random collection of co-orbiting rocks can have all kinds of jolly fun a bumpin' and a scrapin'. Planetary accretion can happen because not all collisions are high-velocity catastrophes that vaporize the participants.

Enough of that rant. But apropos the question of gravity, one thing I find fascinating about the full-globe Charon images is the high relief of the terrain and the jaggedness of the horizon. It's a small body and it's visibly lumpy. Its own gravity hasn't smoothed out all its proud scars. You can't think of Charon as a "billiard ball".

And two large co-orbiting bodies with matching scars? Oh yes I think your theory is plausible, podrock. You may be on to something.