Speed regulation testing used to be defined by ASME standards; I think IEEE is the current standard. I am familiar with the procedures for steam turbines (ASME-PT20.X), but I would think those for diesel s would be the same concept. The concept basically is, with the governor in “droop” mode operation, the “Speed Reference” is set to “Rated speed” at “Full load”. Then, load is reduced and speed allowed increasing to the “No load speed”. Regulation would be the percent change in speed from “rated speed full load” to “speed no load”. For example, running 1800 rpms at full load, speed would increase to 1890 rpms at no load with 5% regulation. (1890-1800)/1800X100.
With steam turbines, the test can be performed from “HSS Speed, no load” to “ Speed, CVWO” by isolating the steam supply. For example, running 3852 rpms at HSS, the speed would decrease to 3672 rpms when the CV reach wide open, with 5% regulation. (3852-3672)/3600X100.
Some governors can have “isochronize” mode. This mode will sense there has been a “droop” and will automatically change “speed reference” to bring speed back to rated (with in a set amount of time) I think this may be "island mode". In which case, besides the regulation droop, there will be a time constaint setting. The very few isochronize control systems I have seen, the regulation was set in "grid mode" and the time constaint set in "island mode". They were electronic and calibrated with simuated speed.
Speed Regulation is a test that I routinely perform on turbine governors. Besides the regulation of load as frequency varies, it is important to verify that the governor will respond properly to prevent dangerous over speed in the event of a load rejection.