There seems to be an incompatibility in the results for average illuminance between Agi32 and Licaso run on the same day and hour. AGi32 reports a range of Maximum=47917 lx Minimum=168.0 lx and Licaso Maximum=328 lx Minimum=90 lx. Both use the same daylight calculation parameters.
I tried checking the calculation method, the north, sky condition, and date and time. There shouldn't exist such big difference if licaso is a AGi32 "plugin".
Did anybody have the same issue?
Once the model was corrected for correct window orientation (Daylight Transition Surfaces) the results are as expected. Please see this article regarding a comparison between Licaso and AGi32.