Admittedly I haven't researched it that well, but I'm curious how often they would have to aim their sensors at the sun to receive from a neighboring satellite. I'd have to assume those links would need to be subtracted from the system, which could slightly drive up latency around sun up and sun down for a part of the planet.
The sun is pretty small compared to space, something like .0005% of the full sky sphere, and almost half that time it's covered by Earth. Since there's no atmospheric glare in space, I suspect it's not a problem given decent engineering.
Given communication is point-to-point, and very coherent, I'd think this is extremely easy to solve if glare in the system was a real issue. Aim the receiver through a tube (or array of tubes) painted black, for instance. Certainly easier than building an equally precise telescope.
