HOMER uses a very simple model of the PV array. It assumes the PV power output is linearly related to the global radiation incident on the panel, regardless of temperature. A panel rated at 5 kW puts out 5 kW at an incident radiation of 1 kWh/m2. HOMER assumes that it will put out 2.5 kW at an incident radiation of 0.5 kWh/m2. I know that thin film PV has a lower efficiency than polycrystalline silicon, but that is no problem for HOMER. You enter the rated power of the panel, and if a 5 kW thin-film array is larger than a 5 kW polycrystalline array, that is of no importance to HOMER.