They wouldn't have to constantly give us new rules or material to keep us interested. How about giving the playtesters some goals instead? There must some things that they still need more info on. Once a week or two they should propose a new test. Examples -
Everybody fight room XX and time your combat, we want to see exactly how long they are taking in real time and number of rounds.
Everybody run an all thief party. Report experiences.
Gladiator fights! Everbody fight pc vs. pc. Who won?
etc.
There are two groups who could run this approach: developers and manrketing.
Developers are constrained by the amount of analysis that can be performed.
Coming up with tests that don't cover the same territory as being currently worked on can be challenging. They would swmp themsleves with data and the need to revew the reported experiences, determine if they fit with expectation, and determine if further development as necesssary or if the tests are no longer relevant due to other changes already in the pipleine.
Marketing doesn't care about the analysis. Its goal is to maintain interest and excitement. But the challenges and responses gathered by marketing are unlikely to feed into development changes for the reasons given above. The big downside from marketing would be once the testers realised their input wasn't being critically examined the test pool would shrink considerably and the public outcry over the 'useless' testing would outweigh any advantage from trying to keep the pool engaged.