Let's take a specific example: the concept of "agency costs". An economist defines these to be the inefficiencies resulting from one's agents (those doing work for him, and those doing work for those agents, etc.) having different incentives than himself. MBA program include agency costs as one of many economic concepts discussed.
A friend with a Harvard MBA was selling his house while living in another city. He noted that the traditional real estate agent's "6/4" commission structure (6% of the first 100K and 4% thereafter) causes a rational agent to sell as many homes as possible without regard to the actual selling prices. Why bother seeking out a buyer willing to pay $20K more when you only get $800 of that? It would be easier to seek out an additional seller instead. An agent incentivized with this structure likely would not act in his best interests.
So, my friend persuaded an agent to accept a very different plan: 2% up to the price he felt he could get by putting out a FSBO sign in the front yard and 10% thereafter. With this, the agent could earn significantly more by working to find a high-paying buyer. The house ended up selling for a little more than the "FSBO sign" price, so he at least saved on commission.
Those here without MBAs might also think in this way -- much as those without CS degrees might also understand algorithmic complexity. It's not that the content of a top MBA program can't be learned through self study in the same way that good developers are sometimes self taught. But, to suggest that the fundamental content of an MBA program is "bullshit" ignores what is valuable.
Perhaps we are most aware of frat-boy MBAs' self-important yapping, but the fundamental concepts of these programs are generally sound. I have an MBA from a less-than-Harvard program and have found what I learned there to be of benefit. And, I consider myself pretty much bullshit free.