After finding the issue about dz2 orbitals (of the sample) having wrong intensity (by factor 4 if I remember correctly), and repairing the issue. The intensity of dI/dV I and IETS in the FePc example was lowered by a lot (and not matching with the experiment anymore).
New DFT calculations are on the way. I will try to re-run those PPSTM calculations then.