My main point was to be experiment driven and let the requirements flow from that. The idea of filtering by early adopters is just a suggestion on how to think

My main point was to be experiment driven and let the requirements flow from that. The idea of filtering by early adopters is just a suggestion on how to think "experimental results first" and that you don't need to run your experiment across 'everybody' in order to get the results you need (as there are surely some people in this set who demand polish/perfection). In regards to what you are saying about early adopters giving you false positives, it depends on what assumptions you are testing in your experiment. What is your hypothesis in this experiment?