It’s probably quite easy to push Moore’s law in the other direction as you suggest. But more effective handhelds would also have to be more-usable, lighter and energy efficient. This would mean voice-recognition software (for illiterate consumers), lighter screens and alternate-powered PDAs (Solar, human-powered) for places where electricity is a problem.
All this would mean more expensive PDAs to begin with but with low total cost of ownership (TCO). Which need not be a problem as long as potential consumers have access to credit to buy such a thing.

Ok, I agree with you on the voice-recognition part. But distribution and maintainance are at least as important as the cost of manufacturing the device and we need to tackle those during requirement analysis.

Another important thing is the software. What kind of software would be relevant for a guy (hopefully more women than guys) earning $1 a day? Money that otherwise could go for a girl-child’s education.

Spreadsheets and Schedulers? Unlikely. Maybe some sort generic communication software that tells her about the market for her produce, helps her communicate with relatives in the city, and help form a political action group against the big dam thats going to flood the village.

But is that something that cannot be acheived by regular GSM cellphones and SMSes? Is our desire then to create a $10 cellphone? If it is that, its reasonable enough. But with $50 cellphones available already, is this idea that revolutionary?

PS: I’m sorry about the delay with the reply, I cannot access the site from work.

PS2: Also html tags dont work when I post :(