I first noticed the tendency of product revisions to destroy power users' hard won knowledge back in the last century, when Microsoft was the prime offender. This was my first experience of "products that only care about unboxing, not deep dive". This is also about the time when I first noticed new hires in software engineering teams refusing to learn to use the "old tech" already in use, and insisting that we all move to the latest and greatest; at the time, this mostly involved documentation, which progressively needed to be Word, html, wiki, and (currently) MarkDown so as to satisfy the new grads. (The net result was chaos, with older docs not editable - or, sometimes, locatable - by newer engineers. We usually had at least 3 repositories of still relevant docs in different formats.) But who needs to know what the original author had in mind, if they can jump to inaccurate conclusions and thereby introduce bugs? Later, there was the phase when they all used a Microsoft IDE to edit source code that would be built - and source controlled - on *nix - introducing giant whitespace changes because they failed to notice they're preferred editing environment saved source files in MS$ format by default, and farthermore used a different default number of spaces per tab. I'm not against improvement, but gratuitous change for the sake of change seems to me to be somewhere between extremely selfish and moderately idiotic. And when the new new thing is worse than its predecessor (e.g. markdown, with multiple incompatible "standards"), I try to avoid being swept along.