You probably just need a few prerequisites. Even if you did study data structures, algorithms, and combinatorics previously, you would need to have your head wrapped around the subjects to fully understand this lecture.

I would also keep in mind that Donald Knuth probably spent more than an hour before understanding the topic.

Something that I think anyone can take away from this lecture is a story of discovering new mathematical relationships. He starts with a function that correlates to a logical problem, changes the inputs in an new (strange) way, then studies the output to form a theory that connects to the original problem.

The key was recognizing old results in a new problem.