The PolishedAl function has different energy ranges for the REFLECTIVITY property and for the different components. I think this is inconsistent and should be fixed. Also, the reflectivity is defined up to 4.2 eV. If this material is exposed to VUV light, the behaviour could be wrong.
I noticed that currently these properties are not used anywhere in the code, but they may be used in someone's branch, so I think this could be relevant.