That's an interesting interpretation of the data. I would have read it as, four years well spent developing a mind that can do engineering work. Here are some other points that spring to mind.
First, 'knowledge' is more than the acquisition of facts. This is why we have spent so much time reforming our education system, instead of forcing all students to have a working knowledge of ancient Greek. It may not be perfect, but at least students are currently encouraged to develop their knowledge in a focussed environment, and with experimentation, instead of memorizing knowledge by rote.
Second, university is meant to provide a well-rounded education. You may not see the value of electives in the liberal arts, but consider: do you really want engineering to be reduced to the level of a trade school? Engineers should, ideally, be well-spoken, capable of crafting an understandable phrase, and able to work in a team. This may not be easily achievable with self-directed, isolated studies.
Finally, there is a certain amount of value in having a professionally built curriculum. I will be the first to admit that there are too many commercial fingers in the academic pie, but when it comes to designing buildings or industrial projects, it seems like employing a self-taught wunderkind - or someone who merely thinks that they are - is a little risky.
Efficiency in education is hard to quantify. It's more than simply volume or accessibility of information; it's about applying it meaningfully and developing enough real life experience to innovate in a useful fashion.