"worse" was a big improvement over 4:1 and was acceptable on all but the most stubborn

On hard joints, or on joints where most or all of the bolts are tightened simultaneously, scatter in the ±5-10% range is probably the accepted norm, because interactions are fewer or are nonexistent. Even here, however, the final variations are determined more by the nature of the joint and by the tools used to generate bolt loads than by the accuracy of the ultrasonic measurements.