That seems obvious but it's not necessarily true. It seems to be an increasingly popular philosophy in this data-driven era of telemetry though, so I'd like to

That seems obvious but it's not necessarily true. It seems to be an increasingly popular philosophy in this data-driven era of telemetry though, so I'd like to offer a counterpoint. I want Wikipedia to be as convenient as possible for the editors/authors, even if most users are readers. The authors -- even as a small minority of users -- create the content that, in turn, attracts the readers. It's possible that alienating authors to prioritize the reader experience ends up driving away readers through the second-order effect of driving away authors. (I'm not saying that these particular Wikimedia changes are an example of this, just illustrating a mechanism by which prioritizing the experience of the majority of users can end up harming them in the long run, because second order effects are important too. Wikimedia seems to have thought carefully here, which is good). An extreme example of this philosophy would be noticing that the vast majority of your users never touch the edit button, and so getting rid of it entirely to streamline the interface for everyone else.