It's small, but it's a start. According to the UN's IRIN, about 100 homes near Islamabad are about to be converted over to solar power to test a new model for supplying electricity to outlying communities. Pakistan's goal is to have 10% of national electricity generation come from alternative sources by 2010.
The bulk of the article discusses the various ways in which the Pakistani government is supporting that move to alternatives -- with lots of "planned" and "soon" -- but the real key piece is the final paragraph:
Once electricity was supplied to villages and communities in areas outside the reach of grid-based power providers, a new social phenomenon would be witnessed, Hamid maintained. "Even attitudes would change once electricity reached a village or community in an area where there had been none previously," he said.
I suspect that we'll see more of this, over time. It may be easier to create a sustainable advanced energy (or information) infrastructure in areas without existing legacy/incumbent systems. Introducing new systems in areas with existing systems means having to pay the costs of converting on top of whatever the new system itself costs. Introducing new systems in regions without them can actually be less expensive than bringing them into "more advanced" markets. As a result, previously less-developed areas can "leapfrog" the established regions, a process noted most famously in 1962 by Alexander Gerschenkron in Economic Backwardness in Historical Perspective.
This may be why Linux (for example) is taking off so profoundly in the developing world. There are fewer organizations grappling with the sunk costs of Windows, more new markets taking a fresh look at which solutions work best. In due time, Linux will be the standard approach for the South (the "BrInSA" trinity -- Brazil, India, South Africa), and any attempt to move off of that standard will face its own conversion costs.