I think it's difficult to solve for the surface potential in the face of strong forward Vbs.  Note in that case the b-s diode is forward-biased and will generally carry significantly more current than the channel, so minor discrepancies in the surface potential and channel current will be hard if not impossible to detect.