HOMER uses that variable only to calculate the PV cell temperature, and the cell temperature has only a slight effect on the PV power output. So it makes sense that the PV energy production changes only slightly with changes to that efficiency input.

If HOMER made you enter the PV array size in m2, plus the PV efficiency, then you would expect the efficiency to make a big difference to the PV array energy output. But HOMER does not make you enter the surface area of the PV array. Instead, it makes you enter the rated power capacity of the PV array. So in the simple case where you are not modeling the effect of the cell temperature, HOMER does not even ask you for the efficiency. It does not need that information; it knows that at standard test conditions your PV array puts out 240 W, and by assuming the PV power varies linearly with the incident radiation, it can very simply calculate its output under different conditions.

If you do model the effects of the PV cell temperature, HOMER needs to calculate the cell temperature, and to do so it needs to know the efficiency of the panel, so it can calculate how much of the solar energy striking the panel gets converted into electricity and how much gets converted into thermal energy that heats the panel. For more information about how HOMER calculates the cell temperature, look up ‘cell temperature’ in the index of the Help file.