I think your test is inadequate. To make a valid comparison you should fill an island or at least the same landarea with CxxxxC and with CxxC. The test just shows that in your scenario at the same city count CxxxxC is superior, but how does it look like when CxxC caught up settling the same area as CxxxxC ? You most likely unintentionally tweaked the scenario in the favour of CxxxxC. Do not get me wrong I am no advocate of any pattern and I do not know what will be better. It would be interesting to see whether the higher city count of CxxC would make up the bigger cities of CxxxxC.