Axel,
Do you have any plan to automate the diffusion (both self- and impurity) aspects within ATAT? I mean, in near future? It seems to me that established models, five-frequency, seven-frequency, nine-frequency models for bcc, fcc and hcp lattices could be automated. They all involve supercell calculations, and then extract necessary info yielding Do and Q. Of course, it depends on you and resources you have.
Gautam 