Eventually the idea is to use G4HepEm in Athena by adding it as a new physics list. In the meanwhile it is just a demonstrator that there's something weird going on when trying to add a new physics list at all.
Note the comments - I have two different options for duplicating an existing physics list. The FTFP_BERT_ATL option works fine, whereas the FTFP_BERT_ATL_EMY option does not. A segfault occurs immediately after the EMY list is constructed.
I note that the EMY list is a class within Athena, whereas FTFP_BERT_ATL is part of Geant4 and thus compiled externally.