This is fun, but it's not really a mystery. Your mind is making the connections. This will work with almost any music/imagery combination, although DSOTM/OZ is particuarly entertaining.
This is a variation of what is called the Kuleshov Effect. Kuleshov was an early Russian filmmaker who tried a series of experiments in which he would show a scene, like a bowl of soup, or a smiling child, or a dead body, and then show an actor reacting to it. He then would ask the audience to say which emotion the actor was displaying, and the audience would respond "hungry" for the soup, "sad" for the dead body, etc. The actor, however, had the same neutral expression on his face for each take. He had not actually seen the footage that the audience did. The mind takes a series of codes and cues and makes sense of it within the paradigm of its experience, assuming that two elements in close proximity are related to each other, whether they are or not. Try it with any music and any video. For example, put on an uptempo jazz number and watch commercials on TV; you'll see exactly what I'm talking about.
