It'd be interesting to see if they're sensitive enough to do the opposite detect the gravity anomaly that is the sub. There's a couple near future submarine warfare books where that's one of the devices used to attempt to find subs.
Gravity force is proportional to mass and the average density of submarine is close to water density so (possibly slightly less as positive bouyancy is preferable for safety reasons) so gravity force produced by submarine should be very close to the equivalent volume of water. So instead of detecting big chunk of metal (which is still relatively small for the scale of gravity), sensor would have to be sensitive enough to detect mass distribution within submarine or the minor density difference due to slight positive bouyancy.
I love when HN commenters get out of water on topics. The GRACE retreivals to reconstruct mass changes for "mascons" that are several hundred kilometers in width\height take a month of static overpasses and incredible mathematics and signal processing. There is absolutely no way that their data could be used to locate a submarines, ever.
The main one I've read is Joe Buff's Deep Sound Channel series, found it after a recommendation from someone else on HN even so glad to continue the chain. Fell off it a bit in book 4 but that might have been me burning out on them, I ripped through the first 3 and focused too hard on them.