I noticed that the cellFlux unit normalization is different when using a box mesh and a box probe. When filling the 3d-mesh histogram, the bin entries are correctly divided by the unit (in case of cellFlux 1e-2 to go to cm^-2). Using a scoring probe, when filling the flux into the spectrum histogram (binned in particle energy) this unit factor is not taken into account:
Also the other available flux scorers seem not to use the unit factor when filling the 1d-histogram.
Is this a bug or a feature?