One ship can get the bearings just fine and more ships would just muddle the acoustical noise. The way one ship gets multiple bearings is pretty easy and uses moving bearings. It's a straight forward vector geometry procedure using direction of the ship, relative bearing of the acoustic fix, time and speed.
Underwater, there are several parameters that introduce variability as the sound can bounce around, it's not just a straight line. In addition, I have no idea how fine this towed sound receiver can resolve the apparent direction of a signal or for that matter, how fine the location of the detector can be fixed. The detector is being towed at the end of a mile or two of cable and would be buffeted around by currents and such.
Also with bearings, there is no such thing as an exact fix, it is a probable location with some of amount of error. For example, with two bearings there is an apparent location where the two bearing lines cross but the amount of error is unknown. However, it is more accurate to have three bearings and when you plot these, you have three intersections and when you connect the dots, you have an area enclosed by a triangle with the apparent position somewhere inside the triangle. I am guessing that the reduction in the search area is based on multiple bearings that are enclosing a box bounded by this new and smaller search area. If the pinger will keep pinging, the box will keep getting smaller.
Thanks for the explanation. Is this is the area they believe is 10K ft. plus deep? No doubt there can be distortions, false echo’s, etc especially if there is any terrain other than flat. When I was a kid living out in the boonies in the ridges we’d get a TV signal usually bounced off a nearby ridge.