Some friends of ours have posted a video demonstrating their use of cellular repeaters to bring coverage some remote parts of Panama. Available here: http://vimeo.com/78562876
I think this is a great idea technically. Many many areas in the world are "fringe" areas with spotty or intermittent coverage. Bringing a whole new networks (like our Village Base Station: tier.cs.berkeley.edu/drupal/node/200) could be overkill. Repeaters are relatively simple and inexpensive (~1500USD) and just repeat existing cellular signals, bringing those cellular networks into the fringe areas.
The primary downside of this style of connectivity is that it doesn't have any capability for local customization or opportunity, as it just repeats existing signal. Money goes to the major telecom and, essentially, the local operator is just extending the coverage of the major networks for them. One could argue this would cause the system to be unsustainable in the long term, though the Panama network has been running for three months without any major issues. One could also argue that the telecoms should pay local operators of these repeaters, though we personally wouldn't expect that to happen, given our experience with operators.
As an alternative, it *should* be possible to actually create a repeater that "resells" the existing cellular coverage. If you attach a repeater to an autonomous network, it theoretically could connect to the major telecom while still allowing for local service and revenue. That's something I plan to look at in the near future.