What would be the point of going much more high tech than what's already commercially available? You can buy commercial satellite footage at a native resolution of 25cm, optionally upscaled to 15cm, for $20/km^2. [1] A bit of processing and you could certainly do high precision geographic mapping. In either case that's already enough to track individual people from space.
I suppose even higher and you could start resolving things like license plate numbers from space, but it just seems like sharply diminishing returns in general.
I was referring to radar for building a point-cloud or three dimensional digital elevation model/digital surface model of the terrain, buildings, vehicles, trees, not photogrammetry or direct overhead visual spectrum or IR imagery.
Different thing. But in answer to your question, revisit time and immediate availability of data would be a major thing. The size/scale of the commercial earth observation satellites that have been launched are quite small and likely nowhere near as capable as even what 'black' programs had 20 years ago. Or even having the capability to get near-realtime live visual spectrum imagery for military command/control purposes (e.g: Is that long boxy vehicle rolling out of a warehouse right now a regular semi truck or a transporter-erector-launcher?).
I suppose even higher and you could start resolving things like license plate numbers from space, but it just seems like sharply diminishing returns in general.
[1] - https://xrtechgroup.com/who-has-the-highest-resolution-satel...