I'm not a wireless expert, but then wouldn't it also be theoretically possible to have a network of direction finders? Isn't direction finding also a repeatable set of steps that can benefit from automation?
Yes, though DF can be much more efficient with directional (yagi, parabolic, horn) antennas. If fully automated by network the antennas connected to the spectrum analyzers need to be on two axis motorized platforms.
It would be easier to, you know, secure wireless communications to begin with. It's not like the Feds couldn't arrange to have stingrays that are properly keyed. (And there's always CALEA.) Yes, I know, it would only be easier for new kit, but it will take a long time to get it deployed. But every year we delay this makes the pain worse.
purely Rx phased arrays, unless very large, do not have nearly the gain (in dBi) of a good parabolic or horn. and not nearly as much directional discrimination as a good sized (90cm) parabolic.
in a phased array that is also a Tx this can be partially compensated for with higher dBm output power from the radio itself, but that's not the usage scenario we're talking about here.