No one will pay you to take your sweet time for several decades and publish if and when you are ready to revolutionize a field. You have to get enough done on a short timescale (a few years for a project that produces a handful of papers is fine) for funded work. Most of us also do have our unfunded pet projects that simmer on the back burner for longer, but these don't get anything like full time effort.
And this is mostly as it should be. You write software with short functions and tests, commit by commit, and release and revise episodically. Releasing one enormous, monolithic product after working in secret for 20 years produces science as awful as it does software, for the same reasons.
Just as a side-note: When I started my Ph.D. studies we had an "introduction course", going through requirements for graduating, seminars by more experienced scientists, what to expect and so on. In that course I encountered the absolute worst seminar speaker when it comes to the subject of motivation. The seminar speaker said that "If your research does not have a potential to receive a Nobel prize, stop doing it.". The speaker was at the time in industry because achieving wealth was a similarly admirable goal. It was extremely discouraging.
So, not typical (I hope), but some argues that fame and wealth are the only reasons to do science.
However, I don't think the author's premise is well founded. Every scientist dreams of publishing a beautiful, fully formed scientific discovery that emerges like Athena, immediately rocking their field and changing the destiny of mankind.
Most science is nothing like that. Most scientific results are comprised of unclear results in a specific model system that may have only speculative relevance to the real world (or to problems that people care about). Presentation, and public consideration and public discussion helps to refine models and cross-pollinate ideas that can lead to major discoveries. Little is gained by keeping data absolutely secret until a scientist is ready to publish.