I think the issue is they try to rely solely on data and metrics, which (as in the case of an IP such as the Simpsons) can lead them down a path that further dilutes the brand.
It's also often unrelated to why people go to theme parks and hotels in the first place.
Choosing to put on the TV to distract the kids, or as background noise while you do chores, is not the same thing as spending thousands of dollars to spend a week on a vacation where all sorts of human needs and desires are to be met. Someone who enjoys listening to a particular podcast doesn't necessarily mean they want to spend $150 and go on PODCAST: The Ride 3D.
Theme parks and hotels are their own medium. They are not TV shows, or sporting events, or plush toys. They have their own production, operation and creative hurdles and have an entirely different set of expectations by the customer who choses to participate. They also have their own unique potential to deliver experiences unlike anything else, which is why people travel great lengths and spend so much money to experience them in person.
Recognizable iconography can be a useful marketing tool, or create a sense of familiarity to the audience, but they are not the only reasons people go to theme parks, and they are not a substitute for quality, capacity and service. If a ride fails to deliver on its potential, the popularity of the IP won't matter (as we've seen in the past).
Using streaming data as the sole or primary factor to decide what to build in a theme park demonstrates a misunderstanding of what a theme park or resort even is, or what has specifically given Disney's theme parks their enduring appeal beyond a vague connection between the brand and its assets.