The real danger for the big corporation business model is technological/social shifts wiping their business model out, not competition. I saw this when I worked at the Fed, years ago. I worked in a division that had two major financial products. One was 80% of their revenue and very powerful, a nearly billion dollar business. The other (what I worked on) was 20% of the division revenue. In the span of seven or eight years, those numbers flipped. The big line of business basically collapsed due to changes in the financial industry, and the smaller product basically ate all of its revenue and then some. This meant hundreds of layoffs and some very painful realignment at the executive level, including months wasted on a power grab.
Prototypical example: health care insurance claim management systems. You cannot begin to imagine how hard it is to rip one of those out and replace it.
But the broader lesson, in keeping with the article's balance, is that enterprise tends to calcify and optimize existing processes.
This is excellent, if say, you need to scale a process to 10x the volume without wrecking your margin. Or survive a government regulatory audit.
This is terrible if your market upends and you need to make sweeping changes to stay relevant. (viz Kmart and Sear's slow immolation)
Sucky, slow software is serving a purpose. It just usually isn't the most important purpose in a rapidly mutating world.