I suspect the real purpose of the Poseidon nuke torpedo is to avoid the difficulties of detecting subs by over-pressuring an enormous area all at once, like fishing with dynamite

I'm not sure how practical they are for mapping but the sensitivity of some such sensors is so outrageous you can definitely measure well beyond a few meters of distance.

https://en.wikipedia.org/wiki/SQUID

Managing the interference is probably the main issue.

Actually, I see now the Wikipedia entry lists prospecting among the uses, so I guess it's quite possibly not only practical but even practiced.

My understanding of the physics is that SQUIDs only work at distances of max of 50-100mm, not more...

You seem generally very knowledgeable about this environment and I'm by no means an expert in the SQUID, but the way I understand it it's effectively a magnetometer with sensitivity down to about femtotesla scale. Clearly that could in principle pick up the magnetic effects of many objects well beyond a decimeter, even with consideration of the field rapidly declining with distance.

The main problem in using it, as I understand it, is rather that picking up everything the sensor could detect would usually result in an unmanageable deluge of unwanted information, and it's a normal operation context to filter the signal for distance so as to not pick up too much interference from equipment, staff, etc. But personally I don't see any reason it would be impossible to arrange it to band-pass instead of low-pass in that filtering.