The major problem with field testing is the accuracy of the instruments and the methods used to take the measurements - you must remember that most pump maunufactureres test their pumps under very accurate and controlled test conditions, something very unlikely that you can reproduce in the field without considerable expense and time. Hovever, if you wish to field teat you should measure the flow rate at the operating point, establish the head across the pump, ie measure inlet pressure and discharge pressure and establish the speed and the power input to the motor (assuming it is an electric drive).
This information can be converted for an SG 0.5 and compared to the published curve for one point which should be the operating point - any other duty points are not really relevent unless the pump has to operate at a number of points on its curve.
As for tolerance - who knows, it will depend entirely on the accuracy of the field testing.
