The easy criticism is that you, your wife, and your children are very unusual people, hence your experience may not generalize.

But I think there is a more complicated and constructive criticism to be made, related to what commenter jadagul said. While its true that much of what is taught in school is useless, there are lots of things which are useful to know but take effort to learn. Essentially there is some up-front effort for a reward that lasts a lifetime. But kids have a high discount rate, and the gratification of the Aha! is not always quick enough or large enough to motivate them to learn these things on their own. Nor is what is fun, or what kids want to learn, necessarily the most useful stuff for them to learn.

Of course, then we have to look at the problems with the other side, with forcing kids to learn something, and we see that they are huge. The effort:reward ratio is terrible - unmotivated kids are bad learners, they tend to learn for the tests and then forget, they understand things superficially rather than deeply, leading to all those ignorant college frosh.

My conclusion then is not that the conventional model is right, nor that teachers (or parents as teachers) should just help the kids to learn what they want when they want. Rather, that the job of the teacher is to help the kids learn what they want *and* motivate them to be interested in things which will be useful, show them how things that seem boring now will be useful for interesting things later, and overcome that high discount rate. (hat tip: Olstads)

Here's an example - I fully support the idea of learning by/while doing. Knowledge gained in context is much more interesting and memorable. But a kid who wants to do X is not going to see that science lessons Y & Z are related. Nor will they necessarily know how to do X, or how to simplify X into X' that a kid their age can do. If you just let them "control their time" and "learn whatever they find of interest", they may give up on X as too hard and go play computer games or read a book, and they are unlikely to learn Y & Z. But if the parent/teacher, using their much broader and deeper knowledge, takes the child's interest in X, directs it towards an X' that is within their abilities, and uses it to teach them Y & Z - that's the model that seems best to me.

But I could be wrong about whether that would work, and I certainly agree that I learned very little that stuck in public school classes, or even college, unless it was so fun that I would have learned it on my own.

Also, the fact that games and books have some educational value does not mean that they constitute an education - just like the fact that cookies have some nutritional value does not mean that they constitute a meal. You override your kid's food instints because you know they'd mostly choose fluff - so why then do you give their educational instincts for fluff free rein? Yes, education is different in that forced food still has nutrition, while forced education loses most of its value. But does that really justify unconditional surrender? Why not the educational equivalent of the airplane spoonful of food, flying around and making exciting noises so the kid will want to eat it?

Also, I'd like to add to your list of benefits that the internet offers. By making it so much easier to find answers on your own (you don't even need to walk to the library), its great for both teaching researching skills, and offering instant gratification, important for those with short attention spans and high discount rates, like kids. And me.