The braking efficiency reported by the rolling road (during MOT test) appears to depend on how hard you can pull on the handbrake. But this is not measured by the rolling road. To me this makes any percentage value given as a pass or fail nonsense.

It is assumed that the measurement is made with the parking brake fully engaged - i.e. the handle pulled up as far as it will go. If you fail to engage the brake fully you will get a poor result. Remember the objective is to see whether the brake provides adequate stopping power - if you fail it's likely due to the brake cable mechanism being out of adjustment, uneven application of the brakes on the rear wheels, or brake pads are worn.

Thank you for your answers. The point I am trying to make here is that the input braking force provided by the hand brake will vary depending on the tester. Yet the pass or fail figure of 16% (UK) is given as a definite. It would be a much safer and fairer test if a known and fairly consistent device applies the force to the hand brake. Then a much more accurate braking efficiency could be worked out.

If a 6.6” beefy tester just back from his gym workout can easily put on enough braking force to reach the 16% (say 5 clicks) to pass the test, but my petite wife can only manage 2 clicks when she parks on a 1 in 6 hill. Not to labour the point here, but there is no consistency in the applied force, yet we are talking about one the most crucial components of a car. The same argument can be made about service brake (foot brake) also

I guess the point is that if your wife can only manage 2 clicks in an emergency then that is what they should test to. If she needs it it an emergency she won't be able to magically transport a stronger dude into the car to pull it for her.

Yes - they should test to see if the force provided by the weakest drivers
(measure and define this force) applied to the hand brake will supply the required braking force at the wheels. This would allow a meaningful efficiency figure to be produced.