Here's an great article discussing alternatives to the One Laptop per Child program. It also delves into how we determine social impact and how should be measured:
"There appear to be cheaper, more effective ways to improve education in developing nations than the glitzy One Laptop per Child program"