The voltage drops where the resistance is, to the extent one can talk about resistance in a ballistic conductor. At this point, given the complications regarding the different choices of constraints, and the general requirement for very long screening, I honestly don't think the voltage drop can provide more than a qualitative picture, and would only be of real importance in cases where you might question where the voltage drops more heavily (for instance, in a complex, long molecule with more than one possible tunneling barrier).