> These are good points. I don't really understand your point about the Hubble though that is very interesting. Are you saying that the pictures are too low-res?
Yes. Owing to the laws of physics, the size of the optics and the wavelengths of its sensors, it is limited to a minimum feature size on the moon of about 22 meters (which happens at the limit of it's ultraviolet sensor range, 115 nm, not visible light). To see the Apollo lander as a single pixel, it would need to have a primary mirror with an 11 metre diameter, to see footprints it would need one with a diameter of 150-200m metres. And proportionally even bigger than that for longer wavelengths such as visible light.
> Still, in the case of massive destruction of satellite communications, having 50% availability for crucial communications (e.g. continuity of government) etc. isn't ideal but is still something. 2.6 second lag is nothing if you aren't talking about real-time communications. Issuing strategic military orders isn't sensitive to 2.6 seconds of lag.
Or you could just use all the stuff on the ground. We used radio well before we went to space, and if the cable-based stuff (and line-of-sight microwave towers) isn't secure then neither is the even more critical power grid.
If you're willing to give up real-time comms, then we have a lot of bandwidth available for text messages that's currently being spent on TV.
> You can communicate to half the earth at once, you can maybe replace GPS if all the GPS sats are shot down, etc.
Nope. GPS fundamentally requires you can see at least four different satellites at the same moment. Also, they're in geosynchronous orbit not low orbit, there are at present no reported anti-satellite weapons that can get to geostationary orbit, nor would this be likely due to the energy budget needed to get there. Consider that while getting to LEO essentially requires the equivalent of an intercontinental missile, getting to geostationary requires the equivalent of such a missile whose payload is itself another intercontinental missile.
> Your point about large antennas is taken, but for USG installations, I don't doubt they would invest in a few large antennas.
I think you've not quite taken on board what I said.
768 times larger diameter.
Diameter, not area.
If your ground station is 1 meter across, like some satellite dishes I see, one 768 times larger is about the size of The Pentagon building and its surrounding car park.
While I look forward to us being able to build structures that size (and bigger) on the moon, the SpaceX website for Starship is currently listing prices to the moon of $100 million per metric ton.