Hello,
I agree that you can do without the "certification" aspects of the testing in your case, but the data you need to prove that one blade profile is better than another is still pretty high, and all of the data mentioned above will be necessary, in practice. In fact, you may be asking for even more data - such as blade pitching and flapping moments relative to RPM, wind speed, temperature, pressure, density, etc. Probably even a gradient of wind speed from the tip of the top blade to the tip of the bottom blade (where the wind speed is slower) so that 1/REV oscillations in your data can be distinguished from other sources of error such as yaw error. Collecting that data in-situ on the rotor as the turbine is turning is difficult but it has been done.
Here is an example: NREL/TP-500-38550 (October 2005) Small Wind Research report (Bergey Excel 10kW)
There is a lot of airfoil comparison testing on the NREL/Sandia labs websites, all of which can be found with some google searching.
You may prefer to investigate wind turbine performance analysis codes which can solve these problems numerically. I would suggest WTPerf from the NREL or something similar as your starting point.