It's funny how fortuitous timing often dictates outcomes.
The piece says about CP/M:
> The early business microcomputer industry ran on it and essentially nothing else
The users in the "hobbyist" market, which was very distinct from the "commercial" market at that time, probably couldn't afford CP/M anyway. There was between the two markets, at that time, some, but not a great deal of convergence. With the introduction of the cheaper 8088, the commercial and the hobbyist market was starting to converge. Whatever OS ran on that chip and its successors, and whatever was priced reasonably, would win. (that's hindsight, but still true in retrospect)
I believe that even if CP/M-86 were delivered on time, and its price-point wasn't set correctly, there wouldn't have been the same convergence. However, DOS was priced correctly and there was no exclusivity to IBM, and boom!, we were off to the races where hobbyists devs could migrate to commercial development with all their knowledge and experience being relevant; make the money, and do great things.
The episode created a pipeline which still exists today.