In other words if the binding energy measured for the alloys from the XPS shift to higher energy levels due to doping, does that indicate d-band energy levels shift to more negative values/positive values?
1) A general rule of thumb: if electrons get withdrawn from an atom thus resulting in a higher binding energy for any orbital, the other orbitals will usually see the same effect.
2) The d-orbitals are the valence orbitals of all transition metals which means for these you can't make an isolated picture; there won't be the d orbital of one atom and the other but rather a mixed band that has contributions from both of them.
Merge these two points in a way that all is self-consistent and you have what happens.
Normal XPS lets us know about the local bonding environment between two elements. If there is a negative peak shift it suggests one of the elements has higher electron densitiy which might be because of doping or intrinsic defects generated. On the other hand a shift in higher binding energy can confirm a succesful bonding between two elements.
To know more precisely about changes in electronic levels of a compound, valence band-XPS (VB-XPS) is a good technique. It clearly showcases the Valence band position, conduction band position, with respect to band gap of the semiconductor.