Their use of the term is not meant personally I'm sure, but for anyone applying insights that are ten years out of date in a fairly rapidly changing field of software like this, they are hardly going to win awards for being smart are they?

This happens too much with software - look at various open source filesystems that had reliability issues in the early days (which is pretty reasonable), someone writes a piece that highlights this and years later people are still parroting it as current insight - with the downside being that it can crush a project and thus in time make a self fulfilling legacy.

KDE has always been the most bloated in the initial install. However when you actually try to do something you discover that whatever the environment they are all about as big, with KDE being the smallest since they trade initial bloat for in use leanness.