This sounds like a reasonable speculation to me. Drones would be orders of magnitude cheaper, more manuverable, expendable, not as subject to being obscured by cloudcover, able to survey many places at once or swarm over a wide area, etc.
You can tell a lot about what they're making based on its proximity to other resources, the shape of the buildings, the schedule of supply deliveries, etc. Intelligence analysts are a crafty bunch.
Yes but - shipping schedules, raw materials, invoices and hirings will tell you a lot more about that than high resolution sat photos of a building in an office park in a city center
("National Reconnaissance Office Deputy Director Air Force Maj. Gen. Ellen Pawlikowski [...] officiated at the unveiling of the Naval Postgraduate School’s latest high-tech research and teaching acquisition, the agency’s Segmented Mirror Space Telescope. The Naval Postgraduate School is the proud new home for the Segmented Mirror Space Telescope (SMT), designed and developed for the National Reconnaissance Office (NRO) as a technical demonstrator and experimental testbed for cutting-edge space imaging technologies.")
These instruments are making single-mirror telescopes (like the ones described in TFA) obsolete.
My understanding is that drones are preferred for long term/constant monitoring. Spy satellites are not placed in geostationary orbit, so they only have a few minutes of exposure per day to a place of interest as they fly over.
The Global Hawk [1] drone is optimized for high altitude loitering. Operating multiple drones in the area is one way to overcome the problem of running out of fuel.
This sounds like a reasonable speculation to me. Drones would be orders of magnitude cheaper, more manuverable, expendable, not as subject to being obscured by cloudcover, able to survey many places at once or swarm over a wide area, etc.