[jpfed]

It's not just a problem that he's discussed before- it is the central motivation behind his approach of "person-on-the-street interviews".

His view (supported by his grad research) was that the default approach of presenting information allows viewers to gloss over how the video's information differs from their pre-existing mental models. After the video is done, they can go back to their pre-existing models without actually updating with info from the video, because as far as they're concerned, there wasn't conflict between the pre-existing model and the video's info.

His hope in showing interviews with people off the street is to get people to think about how they might answer the interview question. Get people to remember their pre-existing model. Then hit them with the contrast between their pre-existing model and the truth, so the viewer hopefully makes an attempt to incorporate the truth.

Now, this doesn't say anything about the information density of his videos. I don't think they're very information dense. This reply is just about his attempt to increase the viewer's retention of the information that is presented.