Let me give my best anti-science pitch and anyone can decide the merits for themselves. You raise excellent questions and I like the example of the serious article based on rigorous experimentation, so I'll use your idea. If we want to tell a story that is true to the best of our knowledge about bike weight (how much, where on the bike-rider system, rotating or not, etc., with respect to performance), and it's gotta be probably true and useful to most of our audience, then what we learn from the test has to be generalizeable, and that's why we need...