He make a good point about there being no studies proving "10x programmer". I don't think it is something you can prove with a study. You would first have to define the metric you are measuring. Lines of Code per minute? Time it takes to build a basic fizzbuzz, web app, compiler? Standardized test?
Once you defined a measurement people could study for it. Someone who has memorized a gzip function is going to make a compressor 100x better than I am going to from scratch.
There are very real and noticeable differences between a $30k and $90k BMW other than top speed.
There are NOT noticeable differences between a $30k diamond and a $3k moissanite stone. A better analogy would be $100 "professional HDMI cable" and a $10 standard HDMI cable.
Complexity is a good word to generalize. I think there are a lot of specific reasons for these BS jobs. It is much easier for a manager to report that he has increased productivity in his division 10%. Rather than report that he has fired 40% of his employees because they were not really doing anything useful. In the first situation the manager is going to get a pat on the back for improvements. In the second situation his managers are going to want to know how this waste of resources was happening in the first place. Also the manager could be looking at a demotion since he isn't responsible for as many people any more.