The extensibility of Emacs is a major boon. Here an example: I never did real Emacs hacking, but I learned some Lisp stuff over the years, a bit of Clojure, Racket, and Common Lisp each. Now, I have a very experienced senior coworker which was asking for help - there was a feature which was just not working for his flow, he had tried to configure it but without success. Now, I looked into the Emacs library code and that was a pleasant experience, it is very transparent and readable and easy to modify, so that it was not difficult to adapt it to my coworkers's needs.

The extensibility also means that, because there are many many people hacking on Emacs, that it stays very up-to-date. For example, Magit which is, I think, the best git front-end in existence. Or there is ranger mode, (ranger is a console file manger). In this sense, Emacs is not "antiquated", as the blog article insinuates, but it is much more recent and modern than any IDE. Take, as an example, git support: It took Visual Studio about fifteen years to add git support, in part obviously because the vendor company did not think supporting git was helpful to the companies objectives. In comparison to that, Emacs picked up version control options such as subversion or git almost immediately. As a result, Visual Studio users were left for fifteen years without support for the perhaps most important technical advance in programming. You may now argue, that the "modern" IDEs of course do have support for git, but there are surely other things which will be missing, just because it does not fit some companies narrative or marketing strategy. Another brewing revolution is that programming culture is drifting away from C++ and Java, in parts even from OOP, which is not any more the best option for every case, and companies which are invested in these "technologies" (uh, what a word), will again try to stop the clock, in order to squeeze a bit more money out of it.