Hi Jason,
It sounds doable. One minor point is that for ligand, phenix.pdbtools needs restrain file. For ADP statistics, it seems asking a bit too much.
naively it may seem too much. But when you do selection on your model: for example, select certain protein atoms, water, ions, ligands, etc.. the selection machinery needs to know what is what (at the very deep basic level), and obviously the PDB file alone is not enough of source of such information, therefore the library is needed. So if your PDB file contains a ligand that PHENIX "doesn't know", then you need to define it first before requesting any manipulation on it. Your are right, you don't need it for ADP statistics, but in my example I was doing selection on a PDB file, so if you want to repeat anything similar your need a CIF file for your ligand that is strange to PHENIX. Good luck, Pavel.