1) Some participant in the mesh still has to buy transit from a real ISP at location that has good RF connectivity to some part of the mesh (preferably the geog

1) Some participant in the mesh still has to buy transit from a real ISP at location that has good RF connectivity to some part of the mesh (preferably the geographical center). Unless they are operating out of altruism, they will need to charge money, inject ads, steal data, or some combination of the above. 2) If the usage pattern is primarily about internet connectivity, rather than local communication, then the nodes nearest the ISP would quickly become saturated with traffic. 3) If people can modify the software on their radios (and mesh networking is a very popular idea with the Free Software community) then they are Free not to waste their own resources altruistically forwarding traffic for others. You would need to design the incentive structure carefully. Probably some combinations of micro-payments, surge pricing/auction, and bitcoin. Otherwise a simple reputation system (bad "seed ratio"-type metric, no connectivity). 4) The big one. Mesh networking bridges the gap when you have A->B and B->C but not A->C. But if you are in a challenging enough RF environment that you don't have A->C then you probably don't have A->B and B->C either. RF is barely usable (if at all) without line of sight, and putting a repeater in the sky is a rock-solid method of getting line of sight. Hence it is the basis for all wide-area RF communications systems. Mesh improves upon certain arrangements of $35 Motorola FRS handhelds, but not all of them (whereas "repeater in the sky" architectures are unfazed). Peer-to-peer RF is cool when you are in RF-friendly environments doing mostly peer-to-peer traffic (not access to an ISP). It works in the places where it has always worked, where handheld radios without repeaters have always been useful. Protests and disaster recovery efforts are two such examples. Internet connectivity for the masses is a different beast.