[thkru]

The dead satellites sit at roughly 0 deg inclination initially and then the moon causes their inclination to shift. Depending on where the moon is in its orbit relative to the satellite it can pull the argument of perigee (lowest point) in different directions. The end result is that the inclinations are all roughly between 0-20 degrees but they all have different RAAN and argument of perigees.

[ricksunny]

Brilliant. As a result I'm reading up on graveyard orbits and end-of-life perturbations from various celestial sources. Also, this gem, whose longitude values roughly correspond to what I'm seeing:

last line of: https://en.wikipedia.org/wiki/Geosynchronous_orbit#Geostatio... :

"Geostationary satellites will also tend to drift around one of two stable longitudes of 75° and 255° without station keeping.[21]"

It cites SMAD, which on its face is more than good enough for me, but since I don't have a copy handy, it would still be interesting to know why those meridians would represent particularly attractive ones to drift into.

[cfraenkel]

*Dead* satellites are supposed to be boosted into a graveyard orbit so they don't clutter up the neighbourhood for everyone else. Of course if it just fails, and can't be commanded there's not much anyone can do, but that's relatively rare. Most will be drifting at a pretty good clip, +100 km or more above geosynch.