The device was developed by the the Indian Institute of Technology and the Indian Institute of Science as part of a national initiative to provide technology to the educational system. If the R&D has already been done, and that substantial capital expenditure doesn't need to be built into the price, and it'll be using a Free operating system, it's not too far-fetched to claim that the hardware could be built for $35. The iPhone costs less than $180 to make, and it's substantially smaller, more powerful, and more feature-rich.
I think that the best thing that could happen to this project would be to make these devices widely available to enthusiasts in the developed world as soon as possible, at a small profit. At under $50, a lot of geeks would buy them as curiosities, and I'd bet that a robust hacker culture would develop around them. Not only would some of that enthusiasm result in software and hardware improvements that could be folded back into future versions of the $35 device, but the more people out there using these, the more meaningful students' work with them would be, since they would be learning with a device that's popular with hackers worldwide.
On the OS front, I can see dozens of customized Linux distros being released for this device, to serve various purposes, from scientific instruments to media players, which would only make the device's original aim of improving technical education for India's students more easily achievable.