After thinking about it, it might be better to go with non-profit route. Get all books donated (ideally from first world where the donation can be apply to tax deduction). Network with church groups where they can pay for the initial shipping cost to your countries. This solves the cost of content acquisition.
Once you get hold of the books, add them to your market place and allow anyone to pick up the books for free. If someone else wants to read that book, they make a request on the market place then it's the responsibility of the current book holder to decide whoever is on that waitlist. Ideally someone close by and who has higher needs. This solves the problem of shipping cost since it's up to the users to share them amongst themselves. The market place keep track of the supply/demand and can make suggestions.
You can do more complicated interactions when the system has more users. Like if a remote village has need for a list of books and a user is scheduled to pass through that village. The marketplace can facilitate all nearby bookholders who have books on the waitlist for that village to give books to that person travelling.
By cutting out "all" of the transaction cost, you can focus on value added services where you can add margin (profit). For example, allow publishers to give out new books in circulation. Newspapers, political ads from politicians, ads flyers from local businesses etc. stuff that readers might want and don't mind sharing. If users don't want the ads/services, they can trash them. Otherwise, pass them on. Interesting way of measuring success of x campaign. Use the money you make from those services to buy more books and expand network or subsidize the network to reach more communities!
That would be a nonprofit that I won't mind donate to.
