i see - i don't really know anything about gerbes
this might change once i read more -- but - at least - the intuition for a category fibred over groupoids is to think of vector bundles over topological spaces, so your CFG is the cat. of vector bundles F: C --> T = Top, and the axioms defining a CFG makes a lot of sense if you have X-->Y in top, and V in C where F(C) = Y, then there exists an object in C : W--> V , with F(W) = X - which is to say that you can pullback vector bundles
in moduli problems if you have a family of ___ over Y then you can pull this back to a family of ___ ov…