I wonder, since this scenario is something I would totally do if I was organizing the tests, if they are intentionally not sending letters to some people just to see what happens? i.e., How many people click on the links, how they use the system compared to those who have been explicitly told they are testing a new piece of infrastructure, etc. I mean, the more they increase their sample size by adding resorts, the more flexibility they have in doing things like this and analyzing different results.
One example I could think of is this:
Average Joe's family doesn't get a letter inviting them to test, but instead the website just acts as if all of the FP+ and MagicBand business is normal. They order bands and select experiences, and they head to WDW. At the same time, Average John's family DOES receive a letter, and does the same thing. Average Joe may or may not know the big picture of whether or not this is testing like crazy, or if it's just a new thing, or what. John, however, knows. Then, near the end of their respective visits, the families are asked for feedback on the MagicBands and MDE. Joe would (most likely) see anything negative that happened differently, possibly being more likely to share and not make excuses like "it's ok because it's still testing" than John, because he was not told outright by Disney that it was a test prior to his vacation. He probably knows by now that it is testing, but he still would have a different perspective than someone who was asked to participate with full disclosure, if that makes sense.