He needs to do the same experiment during the wintertime when the angle of the sun changes. You won't get accurate results until the experiment is performed year round. Different angles of inclination perform better at various times of year, and there are some thoughts that multi-inclined arrays average the same output as uniform arrays.
Remember, a tree only has leaves in the summertime, not the winter. :)
I may have my logic backwards, it should be tested in the summer time when the sun is higher in the sky. Either way, the test needs to be done year round to see how the results vary.
Remember, a tree only has leaves in the summertime, not the winter. :)