Hi @HudsonLe,
I can’t answer this one with practical experience of the device, but from the technical specification the Sonoptix ECHO’s angular resolution at those ranges is ~2.5° (although the 0.47° beam separation means ~5 beams will be hitting an object the width of that region).
As a rough estimate, if we consider the resolution to be the minimum expected detectable object, it could be as large as \approx \text{range} \times \tan(2.5°) - e.g. ~2.2m wide @ 50m range, ~4.4m wide @ 100m range, and ~8.7m wide at 200m range. If a single beam has a very strong and directed echo response, it could conceivably detect objects a fifth of that size (or smaller, if partial response strengths are considered), but identifying the source of a single bright pixel in a reading would likely be very challenging (unless you know it is the only possible object in the water). I’m unsure how much (if any) of that extra resolution is smeared/smoothed out by the sonar’s receiving hardware and algorithm.
Note that depending on the orientation of the object relative to the sonar beam, and its density difference from the surrounding fluid, it may be possible to detect smaller objects in good conditions, but could be very challenging to detect objects at that size or larger if the conditions are poor. This list breaks down most of the variables involved.