Everyone watch out - I'm looking for you.
With more than a little experience in the matters here, I agree with a lot of the faults mentioned.
However, unmentioned and probably most important, is the difficulty of initial acquisition. The vision presumes the ghost-tracker already has it, but a noisy (pinging) surface ASW sonar is one of the easiest things for a submarine (even a slow one) to avoid as its passive sonar hears the pinger at a vastly greater distance than the pinger can hear its own sonar's echo off the sub. For starters, consider the acoustic two-way travel, then the added attenuation experienced in littoral waters, etc. Given that small, stealthy, diesel-electric/sterling-engined/fuel-celled subs with added anechoic coating are exceptionally difficult to detect in shallow littoral zones when stationary - and doubly so if they have plenty of time to evade - this thing sounds like a non-starter. Also, regarding initial acquisition, they are assuming software as sophisticated as the brainy-ear of a trained and experienced sonarman - ears that are the select of the best and practiced for years in all manner of conditions to achieve competence - unless, of course, they plan a real-time data-link to a sonarman in Gackle, North Dakota.