This post ignores the Lindy effect. Things that have been around for pretty long are likely to stick around for much longer. For example, Unix, Emacs, or Firefox (Netscape) are likely to be quite popular 10 or 20 years down the line. The OS or the editor du jour is not. This is a good rule of thumb for choosing what to invest your time in.
Emacs and Firefox are not popular right now, and the author mentioned using even more niche apps with long history, and a lot of his challenges are with things that have been around breaking other things that have been around, so how does he ignore the effect?
They are a lot smaller % than vs code or chrome, but compared to most other software they have a huge user base. They are popular software.