Personal background: I work on this project, but only as a software engineer -- I didn't travel to Ethiopia, and don't have a development background (other than the last 6+ years at One Laptop Per Child). I'm not speaking for OLPC right now.
> The development community know all about this and are scathing of it.
People taking Approach A often think that the people taking Approach B are foolish and misguided (else they'd work on it themselves); it's not a point that carries normative weight by itself. We have to actually look at the arguments.
> a) these tablets are vastly more expensive than just hiring teachers locally - by orders of magnitude.
Well, steady. The two villages we chose are entirely illiterate -- they have no literate adults, and no-one there has ever been to school. One of the villages has potential access to one school that's 10 miles away and 3000 feet lower, which is not a plausible trip.
So now you've signed up to build two schools, and then find literate teachers who want to live in or spend all of their working time teaching in a remote illiterate village. And you realize that Oromo is highly dialectal, such that the two villages in this case speak dialects that aren't comprehensible to each other; your teacher probably doesn't know their dialect already. Is it still orders of magnitude cheaper than tablets?
> b) More perniciously, the tablets educate in English or Amharic: not in the local language of the population in question, which these children would otherwise speak; their parents are often not fluent in either.
There are no Oromo applications of any kind for Android, so in effect you're saying that this experiment should simply not happen, and these kids should not get to use technology. I don't agree with that.
Even if it's successfully argued that teachers are a better solution for Ethiopia -- I certainly agree that a small class size and a brilliant teacher is a wonderful thing when it's available -- there are a hundred million kids in the world with no access to a school. That seems worth researching a potential amelioration for, right?
Maybe you're missing that this is a pilot-stage experiment that only involved 40 children? It's not being proposed as country-wide policy.