As soon as the analysts making judgments about the evidence are in communication with the prosecutors or police, it's junk science. You wouldn't do a psychology or medical experiment without at least single blinding, and certainly not with the experimenter telling the subjects what result they want!

Look at this damming quote - "ShotSpotter strongly insisted it had not improperly altered any data to favor the police's case". That shows they don't have a system in place to prevent data being altered or to prevent themselves knowing what the police want, so all their findings are suspect.