The branching ratio of all very weak gamma lines (< 1-e5) is artificially set to 1e-5. This causes some rare lines to be produced more often than in the reality. While in most applications the effect is negligible, it may have an importance for  the 2505-keV line of Co-60 (table of isotopes: 2e-8, Geant4: 1e-5).

The problem comes from line #189 of G4NuclearLevelManager.cc:
const G4double minProbability = 0.001;

I suggest this value to be substantially lowered, e.g.
const G4double minProbability = 1e-8;