I really don't want to take all of this point by point, because for the most part, 'Accurate but Incomplete' is, to borrow your assessment, accurate but incomplete. If you don't mind, I'll respond to a few specific comments.
Any software engineer will tell you it's impractical to store information on a "continuous" basis. Discrete collection points have to be established in order to populate a database. "Mr. Smith entered the theme park at 9 AM. Mr Smith entered the Peter Pan queue at 9:30 AM. etc." We could guess where Disney intends to collect such information but, from an implementation perspective, this information must be gathered at discrete times and discrete locations.
Whether it is practical or not depends largely on how much money the customer wants to spend, which goes back to how they're going to use it. Your example is good and ties into discrete locations, which is fairly economical and leads to information that is useful in a number of interesting ways.
The cards cannot be read by today's long-range readers. But they can be detected by short-range readers that will be deployed at numerous data collection points. Disney's ability to collect "location information" is limited only by how many short-range readers they decide to deploy and current technology. As the technology improves and as Disney identifies needs, Disney reasonably can be expected to upgrade its technology and deploy more "location information" readers.
"location information" will be collected for MagicBands. However, there is nothing special about the RFID cards that prevent less-accurate "location information" to be gathered as well. And, as already mentioned, as technology improves and if the business need arises, it's reasonable to expect Disney to improve the "location information" it collects.
There's a balancing act, and while you're correct to point out that that balance can change over time, there are some mitigating factors to keep in mind. For one, the bands themselves are a limitation -- once they're deployed (and especially if they get any traction with folks customizing the bands), they'll be difficult to upgrade, so they're stuck with the first generation tech. If Disney was the type of company to plan ahead and put extra money into things, then I might be willing to believe that they've built extra tech into the bands as a way of future proofing them (if they were that type of company, I don't believe we'd be discussing this).
Another mitigating factor is the utility of higher resolution data. One of the fundamental things that makes all of the modelling work is that it requires working with large numbers. If you are gathering information at a very high resolution, then the numbers of guests in any particular location becomes smaller and therefore less reliable. This isn't a function of the technology -- this is a function of the underlying math.
Look at it this way -- there's a reason that television ratings, political predictions, etc, work with fairly large demographic buckets, and it isn't just for time or convenience. For example, you could measure TV ratings by each discrete year of age of the viewer, but it turns out that the distinction between viewers who are 25, 26, and 27 isn't really meaningful -- it's essentially statistical noise. Gathering location information for guests passing through a large number of small areas has the same weakness -- the distinctions between adjacent areas yields far more noise than useful information.
Finally, remember that the real value in tracking guests location is in aggregation & correlation. Actual paths that guests take aren't something that can be usefully combined or compared. For the uses that require information that is current (targeted advertisement comes immediately to mind), they want to target large enough groups that the expected response rate (generally something in the single digits, if it involves spending money) returns enough positive responses to make it worth the effort. They're not going to advertise to everyone passing the water fountain across from the TreeHouse -- they're going to advertise to everyone between the Crystal Palace & Pecos Bills.
For all other uses, there's almost no interest in the individual data points. How much time I spent in the Emporium isn't really interesting to them, except that they can combine that number with all the other numbers for men in my age group, income bracket, home area, etc, and that they can observe how that aggregate number changes over time as they try different things. (I would like to think that one of the things they'll eventually try is to stock the shelves with a wide variety of interesting, good quality merchandise. although I expect they'll exhaust every other alternative first.)
The company won't have an interest in stalking you - there is no return in doing so. The company is interested in recording your patterns and usages to correlate them to their services and sales.
