First of all, fantastic post. This really highlights an increasing problem in which product developers are focused more on product usage analytics than they are each user's personal experience with the product.

I think there is another way of looking at your Phase 2 though. Your goal is to make your product so good that you can't imagine using anything else. If your product is really that good, I'm not sure that consistent reinforcement is actually necessary. Rather, your product should be so good that users stop using it consciously.

I think one of the best examples of this is Google's search engine. The product itself is so good, and has proved itself so consistently over the years, that most people don't think of using anything else for search. In fact, most people don't think of "searching" at all. I can't count the number of times that I end up on a Google search results page, not realizing that I had gone through the mental thought process of "I need to search for something. I'm going to go to Google, execute a search query, and find what I'm looking for in the results."

Going from thought to results without any significant reinforcements from the product (other than it simply being a great product) is in my opinion, the true sign of a user experience done right.