That doesn't make sense, how can you gain extra detail if the bitrate is the same on each pass?
In CBR encoding more passes will result in a more constant and stabler bit allocation for GOPs. Also the bit allocation for each picture in a GOP will be modified to achieve higher image quality. (CCE User Guide page 35: Multipass CBR)
Everyone who I discuss this with doesn't believe me but maybe in this thread it can be cleared once and for all. Multipass CBR gives a better image quality then single pass CBR. A balance has to be found for the bit allocation between simple and complicated parts in a single frame. But this is the same in VBR, not only is the ammount of bits available per frame optimized but also the bit allocation per frame.