not exactly, I was wondering what would happen if you had three GH antennas back to back to back, each of the three reflectors had their own drive element just pointed in a different direction (i.e. 120^o from the other)

Using your NEC file for the circular GH, this is kind of what I thought (but the final build of the reflector being a lot simpler and more like the diagram I showed) . The issue is that when I tried running this model, the program crashed (could have a few things that are wrong with it since I have no experience with this stuff). If you look at it from above, I think you would understand what I mean, so here is the NEC