So... They didn't test the iPad / content combo to establish usability / feasibility / usefulness prior to dropping all this cash?
That's speculation. Feasibility is no guarantee of performance.
I read the attached article, and there were two specific complaints cited. The first was security, which is a non-functional requirement; that could well be a failure of the customer to do his homework on requirements but presumably a competent and honest vendor could have done a better job on security. It's often the vendor's job to anticipate customer needs, particularly in projects of the type customers don't necessarily have experience with.
The other complaint is that the curriculum wasn't completely implemented. If the vendor failed to deliver something it agreed to, that's purely the vendor's fault.
Sometimes bad vendors happen to good customers. Bad vendors happen more often to bad customers, but every project involves taking a calculated risk.