Your thinking is correct. Control full throttle RPM with load. Then experiment always at that particular RPM until you have whatever data point you were after. Everything affects everything so it may be that (for example) advancing ignition timing at one RPM adds power but at another RPM reduces power. A rich mixture at one RPM may increase power but reduce power at another RPM. To recover even similar data on consecutive days is difficult. The SAE has some charts that produce a multiplier to account for density temperature and so on. Required to compare results day to day. It takes a long time to collect good data..........
Lynn E. Hanover