Licence costs aren't an inherent advantage of 'digital' distribution over paper, though.
But if we could get to the stage where the licensing costs are are small compared to the production costs, the laptop would do better, because it could contain thousands of books.
An individual student may only need 20-30 books a year, but after about 4th grade each student needs different books. They may not even know what books they need far enough in advance to get it.
If the goal is to live like a 300BC tribesman then they don't need books, they will follow adults and learn what works. Even today for the poor that is 80% of their life: learn from and do what the adults before them did.
Where reading helps is when the kids get interested in something they couldn't know. Should they start no-til farming in their fields - done right it builds the soil, but done wrong and there is a complete crop failure. They need to build a new hut - is there modern construction methods that they could apply to build a better hut? Those are just a few of the questions that they can learn from a book, but there is no reason for everybody in the village to learn that, just one specialist.
Currently working in the EdTech space - I have read figured like ~1.9 million new teachers are needed to teach all the students not receiving an education.
Scale that to the students themselves - if they don't have early digital exposure, they very likely will be excluded from being digital natives.
Add to the mix, needing the same curriculum in many languages. Some of these problems can be assisted digitally.
I'm more optimistic about this than you seem to be. The OLPC XO-1 has a lithium-iron phosphate battery [0], which are only now reaching the end of their design lives. Anecdotally, I have two XO-1 laptops, and one still has hours of battery life, and the other has only a little over one hour.
There is no rotating media. There are no fans. The keyboard is a single piece of silicone rubber. I've dropped one of mine from waist-high onto a vinyl floor, and only a piece of trim plastic popped off. (I put if back on later with a screwdriver.)
It doesn't seem unreasonable to me that they could last 10 years in the field, especially seeing as how mine are ten years old and are nearly pristine.
All these licensing concerns exist in developed economies.
In third world countries like mine (India), the government produces text book content, prints and sells the books at subsidized prices. Stationery items are totally untaxed. If you look at the link I provided, you will realize that the most expensive textbooks in India cost about $2. $100 will buy 5 years worth of books for a student. That's much more useful than a substandard PDA with wifi.
I have travelled to India and bought books there. Amazon India is super affordable too. Not to mention the ongoing book piracy that is rampant in countries including India.
Textbooks for higher education cost a few more than $2 each.
I agree that reading textbooks on a laptop is not ideal. But giving someone 600 books in 1 device has merit, especially when those devices are shared.
I do like the comments in this article pertaining to e-readers. I am a very late convert to Kindles, and they are usable.
In India particularly, devices like the $35 Aakash tablet are also making inroads.
It all, like you said comes down to accessibility of quality content. Digital delivery will continue to have an increased role in that.
If textbooks were cheaper, Khan Academy would not be successful.
Open Educational Resources (OER) hopefully will make some inroads on this but until then, textbooks are generally more expensive.