this depends upon how you define "fair". It might work for comparing different cells within the same technology but not for estimating the final power output under real world conditions. Especially when you have conditions with a high amount of UV (high mountain region) because this is almost totally missing in the halogen spectrum.
You can find a comprehensive discussion in the link I added.
As mentioned by Peter above, all your results accuracies will depend on your set-up, and spectrum is ONE of such parameters (power density, stability, uniformity, temperature, optical reflections and electronic accuracy are others) CIGS devices have a lower quantum efficiency compared to silicon technologies in the UV region of the spectrum (record cell from Fraunhofer being only ca 20% @ 400nm).
Dear colleagues, can you provide me data on spectral response of the CIGS, I want to compare the light energy, halogen and CIGS to simulate how much energy will be “missed” if I use the halogen lamp.