Like @
wdwmagic said, typically the reason would be about control. You can't control your own destiny when you ride on a platform that is independently supplied by the customer. The property doesn't control the rate of adoption by the customers and doesn't control when Apple or Motorola decide to iterate their products. The lifecycle on these consumer products is relatively short... 2-4 years.
You can accept that kind of 'take what you can get' when you are doing a simple 'value add' service. There, it's just about balancing the cost vs the gain... will enough people use it, will the value the customer gets from it be sufficient to justify the expense, etc. Disney on the other hand tho was looking at a technology where the token would be the foundation used for essential services.. for everyone. For WDW, the penetration rate for customers needs to be near 100%.. (not just who happens to have a phone..) and the investment in specific reader technology is so massive due to the scale of the project you need to be able to garuntee the usable lifecycle. You don't want to replace thousands of readers to find out in June '14 a vendor you don't have influence over just invalidated all your new readers
I think the main takeaway is 'value add' vs 'fundamental technology'. If you are as big as WDW, you can't build your entire operating world on something you have no influence over.
This isn't to say WDW had to build a CLOSED system.. that couldn't use customer provided devices as an alternative token if the customer had it. But it's entirely a different world to do it the other way.. which is 'customer must provide token'
There are also technical benefits to the design that is highly optimized to your application... like power consumption, control over range, cost control, etc. The one thing going for WDW is scale... meaning they can justify dedicated R&D and production design costs because the application of it is so massive and in theory with a sufficiently long life cycle.