I'm curious what the assumptions are. Looking forward to reading the paper. a kilometer-or-so is about what you get on Earth without a lot of sophisticated co

I'm curious what the assumptions are. Looking forward to reading the paper. a kilometer-or-so is about what you get on Earth without a lot of sophisticated corrections, averaging, and kinematics. So, if they're not doing all that stuff, they could be doing quite well. (on the other hand, one of the bigger correction terms-- the ionospheric delay -- they don't have to deal with-- but they have to deal with all of their measurements being in "one direction"). If e.g. they don't know about the moon's relative motion, that's a big disadvantage. If, on the other hand, they get the kilometer after a -loooot- of averaging, that's quite bad. I don't know how big of a fleet you need to make this worthwhile, though. Just one satellite in a different direction would collapse that big error ellipse to a much shorter arc.