HarveyGee - sorry to confuse you - compliance with 20.7.4 is easy if you choose to use the old P Charts. Just read I Fly's post about the letter he has from CASA on how to do it.
It seems that this subject is causing sleepless nights for Gaunty and myself. I've done some more digging on the origin of the P Charts for my airplane. I don't doubt that some actual airfield distance measurements were made on an 8KCAB when the first one arrived in country back in '74. It seems that it had a constant-speed propeller and that would account for the discrepancy I mentioned earlier. There are quite a few with the optional fixed-pitch propeller - I wonder if the Department realized it back then or whether it was simply an oversight that my airplane received the wrong P Charts. Of course, most of us are reasonable people and wouldn't get in a situation where we'd have to consider whether a 400 m strip was long enough.
Gaunty refers to overseas performance data being generated for "sales" purposes as the reason for the Department doing its own testing. Quite true but one slipped through. I spent quite a bit of time with one of the old guys who'd been around the USA GA engineering/manufacturing industry for a long time. In those days the USA AFM did not require airfield distance data. I don't know where he got the data from. Maybe he paced out the ground roll on one occasion. I do know that his rate of climb is about double what anyone will actually see when flying the airplane. He presented his "sales" info to the Australian Department who dutifully applied the 1.15 factor and made P Charts out of it. Some years later the USA FAA wanted airfield distances for a new model of the airplane and he submitted the Australian P Charts - now "approved" data. The FAA rejected it so he then had to conduct some flight tests. It'd be interesting to compare those results with the P Charts.