Are you suggesting that flesh-and-blood teachers or brick-and-mortar schools are less likely to spread misinformation or leave students with misconceptions?
Possibly, because when you leave school you can realize your misconceptions and learn from other people, as well as pointing out other people's misconceptions. But if everyone uses Khan Academy, everyone will end up with the same misconceptions.
Except the self-fixing process you claim to be characteristic of physical institutional learning only happens when there is a significant number of people in the class that are A) In a position to think confidently and critically about the material B) willing to invest time and effort (often lots of both) in finding the correct conclusion and C) in contact with you about their insights. I have found the confluence of those factors to happen exceedingly rarely in physical classes and significantly less rarely in hybrid physical/online classes. This isn't surprising because lectures place severe limitations on two-way communication that online environments don't. With a traditional class structure (lecture/discussion section) I suspect at least 3/4 of questions go unasked and at least 3/4 of the received answers never get shared beyond the immediate group of individuals doing the asking (in comparison to an online forum, where the answers are googleable or at the very least spread to the entire class).
Also, online learning wins in the limiting case where courses have been iterated and feedback / FAQs taken into account. In a brick and mortar institution, "institutional memory" is effectively lost as professors retire or shift interests (if it ever existed in the first place) and their replacements introduce their own slew of idiosyncrasies, poor methodology, and misconceptions. In contrast, an online lecture never suffers this degradation. Sure, new lectures may replace old lectures (presumably at different sites), but both will be available for a time and during that time, students will pick up on inconsistencies. That's not true of brick-and-mortar lectures, where the only source by which a student can verify the material they are presented is the textbook or its alternatives. Combined with the tendency of courses to discourage reading the textbook (it's almost always a less point-efficient use of time to read textbook chapters than to do homework, practice problems, and go to office hours, which provide information better matched to the material that will be tested on), the status quo is a recipe for failure. My experiences confirm that this recipe for failure produces failure (in the pedagogical sense).
The third factor that favors online learning with regard to misconceptions are its twin killer features: the pause button and the rewind button. These reduce self-inflicted misconceptions tenfold.
So, sp322, I think the opposite of what you stated is true. I would welcome a conflicting opinion, because usually when I go on rants similar to the above three paragraphs all I get are nodding heads.
Yes, lets just perpetuate mistakes in perpetuity. And I highly doubt everyone would end up with the same misconceptions. However, the fact that Khan Academy is a) not rigorous, b) has no type of human interface with the teacher, c) has no type of input from professors and lead academic institutions, d) is purported as an educational savior by influential people when there is no such thing, e) uses outdated and confusing teaching techniques, f) has yet to put problems over exercises, and g) has no type of human interface between students it has a long LONG way to go.
Possibly, because when you leave school you can realize your misconceptions and learn from other people, as well as pointing out other people's misconceptions. But if everyone uses Khan Academy, everyone will end up with the same misconceptions.