I remember a paper from the 80s, the first one reporting the localisation of some cytoskeletal protein, vinculin or talin or something, where the authors present a set of microscope images showing that the protein is found exclusively at the ends of stress fibres (the main structural members of many kinds of cell). Great finding! And then they show a picture of a cell where it isn't, and is instead assembled into lots of little rings. The caption says that they have no idea what kind of cell it is, or why it's doing that. I just can't imagine that happening today.
In this case, this is a review paper, so that doesn't directly apply. But i think the ethos still pervades the writing; the authors are writing to explain what's known, not to advance some hypothesis or demonstrate how much they know.