This is somewhat of a loaded question, with lots of room for interpretation. ASSuming we are talking about illuminance from daylight, the simple cop-out answer is "it depends".

eQUEST, and more specifically the engine behind it, can solve the light transport equation accurately, or reliably, in certain cases, and further assuming the input is sound. But both DOE2 and EnergyPlus suffer from geometry models that (can) bear little resemblance to the actual building architecture, and since that building form is affecting the distribution of the daylight, you can see how loads of cases will fail to produce reliable results in this area. This is one reason--there are several others--why the OpenStudio project offers Radiance as a calculation path for the daylighting design of your whole building energy models.