I've got your low-cost device right here...
According to Wikipedia, 73 million people in China access the Web using mobile phones. (I've seen figures as high as 170 million, but that would equal the total number of Internet users in China as of 2007. Seems doubtful.) 73 million is about 30 percent of China's 253 million Internet users. A just-released study by Vital Wave Consulting states manufacturers of low-cost computing devices--sub-notebooks, ultraportables, whatever--are targeting the emerging middle classes in countries such as India, China and Indonesia. There are few devices specifically designed for the majority populations of these countries--rural, poor, off-the-grid, and generally faced with choices that make access to information a luxury.
Why is this? What about the 2 billion or so people at the bottom of the pyramid, don't they comprise a massive market for low-cost computing and Internet access?
I think, perhaps, not.
According to Richard Fuchs of IDRC, mobile Internet access (or using a mobile phone to access the Internet) is growing faster in developing countries than desktop Internet access. (I met a guy installing 3G in Bengal in 2003. I worked on a project using GPRS in Haryana State that same year. We were "mobile Web" before we knew what it meant.)
Putting the next nail in the coffin of consumer-oriented low-cost computing in poor countries, according to Simon Batchelor of Gamos Consulting, the introduction of 1 mobile phone into a village in Africa increases productivity 10x, while the second phone increases productivity <1x. Villagers share information. (Both of these nuggets of information have been shared with me directly, I can't find them on the Web.)
Thus, given the growth of the mobile Web in developing countries, combined with the tendency for the information and communications provided by "first-access" devices to be shared among poor users, well-managed design, manufacturing and distribution of low-cost, low-power computers by commercial entities is not going to target the poorest of the poor, or even the generally poor. Those 2 billion poor people, as a market, can be cut down to maybe 20 million early adopters, because they'll make crucial information available to others in their villages. And even those village-based pioneers of Internet usage are going to--based on cost, based on their living circumstances--opt for mobile-Web devices not tiny laptops.
Hell, I can update my facebook page using a free application specifically designed for my mobile phone. Which would suggest that the device that's going to crack open the Web make life-critical communications available to the world's rural poor is going to be an iClone.