How is it sound? C2 function of AEW&C aside, the distance of satellite is so much greater and its power output so much lower I don't see how it'd be able to replace even a fraction of the function of AEW&C.
You could theoretically cover the entire sky with tens of thousands of satellites that can band together to form a MIMO radar over any specific area of the planet to lower the power required by any specific single satellite, it comes with the added benefit of being able to have simultaneous coverage over the entire planet barring specific locations like directly over the poles for which you could just use normal AEWC aircrafts to plug up. Plus satellite could orbit as low as 180km which isn't insanely far away, Hubble's brothers and sisters orbit in this general region of 180km-250km and so do some other spy satellites. Also planes even stealth aircraft usually have a much higher RCS from top down which might remedy part of the greater average distance involved and lower power output of satellites. But as I said such an approach is extremely time consuming and expensive to setup and may require some currently in development technology such as GaO radar, more efficient energy storage/collection systems and advanced launch vehicles that are capable of lowering the overall cost to increase effectiveness of the system
PS: I looked it up and someone did the math for the max theoretical power generated from the solar panels on an early starlink satellite
"We don't know the exact dimensions of the satellites, but
. I'll assume that diagonal is 4.4m. That gives a satellite length of ~3.1m. From the
, we see that the array is as wide as the satellite, and a pixel count (keep in mind, the view is skewed, so the count is not perfect) shows that the panel is about 3.5 times as long as the satellite is wide, or 10.9m. That gives a solar panel area of about 34 m2 ."
We know that ISS generates 240kW of energy for 2500m^2 of solar panels at approximately the same orbit as starlink satellites, that comes to 0.96kW per 10m^2 of nearly 30 year old solar panels(I did the calculations myself and found that the panels were only ~7 percent efficient a far cry from current panels that could reach up to 25 percent efficiency). So, each starlink satellite should have atleast an average power of ~1.5kW over each orbit including depending on the specific orbit approximately half the time spent in darkness. So, with new solar panel technology and battery technology an average power ~1.5kW for the antenna plus a bit more to operate other critical systems for a total of 2kW average could in theory be doable for a theoretical mass produced spy satellite. 1.5kW average power for the antenna is no way a small amount, FYI that's basically the same average power for a medium sized fighter radar. Now imagine thousands upon thousands of these in low LEO that could also band together other satellites in the constellation to form a MIMO array for better detection range/resolution
