Feasibility of free calls in real installations of Fairwaves/OpenBTS

Recently Steve Song pointed me to the great project Freedom Fone (don't mess it with FreedomPhone, which is a different project). FreedomFone guys wrote an article about the horrible situation at Dadaab, the biggest refugee camp in the world: http://www.freedomfone.org/news/village-fone-dadaab

Steve pointed that VillageTelco may not be the best solution for the communication problems there. My reply is that a combined VillageTelco+Fairwaves/OpenBTS would be a much better solution:

From what I understand, the best solution might be a combined Fairwaves+VillageTelco system. As most people have their mobile phones, it would be beneficial to have a mobile network there. At the same time, there are few stationary places, like schools, agencies officesm etc where MPs could be installed. Our system is great, but it has a very limited capacity — 15 concurrent calls per base station sector. So we should do the best to offload traffic from it, and VillageTelco is the best solution for that.

For the same reason, it's unwise to provide completely free calls on the mobile network — people will quickly overload the network with random chats and more important calls couldn't come through. This is a bit different then with VillageTelco, where you can't have more calls then you have MeshPotatoes installed. Thus I see that if we were to install a network in Dadaab, we had to charge a small fee for mobile-to-mobile calls. While calls from mobile to a set of «emergency» numbers had to be free. E.g. calls to the Freedom Fone services would be free to provide maximum value to the people. MP-to-MP calls could be free as well to allow unlimited calling between important official offices. Then, the gathered money from mobile-to-mobile calls should help to support system maintenance and make the system sustainable.