If you could get some of the big charities on-board, the costs would drop dramatically and add in the possibility of manufacturers/vendors chipping-in by donating stock that now just gathers dust.
If some of the larger charities could work with local people to set up a small production facility to build Raspberry Pi units which could be sold to local residents, they could become self sufficient and bring computing to areas which either don\'t have access or must rely on internet cafés.
Of course, this on its own will not be enough as you would also need to support this with education. Teaching people to use the computer so they can teach others in the area. I can see great potential for the Raspberry Pi, particularly in developing areas where education and entertainment are things that only the wealthier can obtain. There was a big move in Brazil not so long ago to learn and make use of Java because it was free and some good stuff was done (although not enough to make a big difference IMHO and I guess it also depends upon your thoughts on Java generally).
If the government and/or charities can fund initiatives to give every child a laptop, or provide free mobile phones to people, an extremely cheap solution such as the Raspberry Pi should be more achievable, less costly and allow more ways for people to learn and develop.
Looking back on this post I realised that it might be easy to get the wrong idea about what I said about setting up production. What I meant by this was for people to get the completed Raspberry Pi boards and then put them into cases and add other hardware to them.
Thought I would clear that up a little.