This is often said, but the limited evidence available today makes it seem that it's not his ideas that are controversial but his explanations. And the reason those explanations are controversial is because they take a fresh look at what was actually being danced, free from the baggage of much of the misunderstood incumbent technical language. Unfortunately, that language had/has too often been allowed to divert attention from the dancing it was intended to describe - people were trying to literally execute misunderstandings of the descriptions. A good example is the camp of people who attempt to reflect a theoretical difference between CBM & CBMP by oddly constraining their CBM to avoid ending in a position that they, based on misunderstanding of the incumbent language, think would be miss-applied CBMP. They believe that any twist of the body into an opposite side lead relative to the feet is necessarily CBMP, when in fact that is not the case. CBMP of a foot usually occurs as a result of diagonal movement across the feet, and never occurs as a result of body rotation during the step when that foot is placed. And in fact, properly done CBM will in most cases result in an opposite side lead that does loosely resemble CBMP, but by definition is not CBMP. In contrast, Scrivener starts out by acknowledging the similarity, then simply stating that CBM is when you rotate and CBMP is when your step direction is towards the other side of the body.