That chart comparing the handful of prospects that go straight from AA to the majors with prospects that play in AAA is pretty disengenous. Of course the handful of prospects good enough to make that jump are better than the 1-step at a time guys. And it is not like every club in baseball does not try to identify them also. They just don't have more than 1 or 2 of them.
While there are valid points in there about development and advanced analytics, this article also is an example of analytics people sometimes being misguided with their models. One can apply the analytics and new methods while still using what looks like a traditional minor league ladder. I also don't think the 9 versus 7 affiliates means too much as it is correct that the extra 50 prospect one signs to stock those teams are mostly prospect filler. Of course that is true because teams have the same draft pick systems and international limits. Everyone beyond that has little chance of success, though I would argue the marginal few millions of dollars spent developing those players might be better spent there than anywhere else. Houston has less money than the Phillies so it might be the natural decision for them to have fewer affiliates while it is equally correct for the Phillies to spend the extra money.
Articles like this assume organizations can't walk and chew gum at the same time. Baseball is unique in that prospects generally need years of seasoning and there is little in this article that disputes that other than giving advanced analytics instruction to maybe speed up the development of the best prospects and that will happen anyway.