Pulling smoke past the prop won't really tell you much. The airflow is far too slow for the patterns to be applicable to the operating speed. Far better to run the prop on a motor where you can check the rpms and current. Measure the watts in (volts x amps) for each run and the RPM. Keep the RPM constant by varying the voltage and current and chart the amount of watts for each mod. For this to make sense you need to use the same brand and version of prop for all testing. Try stock, dimples, trip threads and anything else you want but make only one change at a time and keep the RPM constant for all tests.

Granted this is a static test and folks will be right in stating that the results will be skewed by this but it would be hard to do this in the air. To reduce the static issue be sure to use the lowest pitch prop you can for the testing.