It sounds like he's confusing the high resolution imagery satellites used for reconnaissance with the wide area surveillance satellites used for launch detection. Those are very different systems that meet very different requirements.
Photographic reconnaissance (or imagery intelligence [IMINT]) involves taking high-resolution images of specific points of interest. This requires a camera capable of high resolution, which almost always forces a narrow field of view. In the 1960s, the primary imagery satellite system was the recently declassified Corona system. Corona was a lot smaller than the Hubble.
Launch detection requires surveying a very wide field of view - typically the entire visible face of the Earth. This necessates much lower resolution (and no, I won't tell you the resolution.) The current example of this type satellite is the DSP. In the 1970s, earlier and smaller versions of DSP did the same mission. Before that, the MIDAS and Vela satellites systems did the mission. Both of those systems were much smaller than the Hubble, too.
IMINT satellites are kept in LEO in order to get the very high resolution needed for reconnaissance. Surveillance satellites like DSP are normally kept out at geosynch to get the very wide field of view necessary for the mission. The planned future launch detection system (a href=http://www.fas.org/spp/military/program/warning/sbir.htm>SBIR (space-based infrared) will consist of both high fliers and low fliers.
Space/Science » in reply to Optical Interferometry
He got that part wrong
