I have a comment on the idea itself, not how it summarizes the book (I forget exactly what the book said about this).
Sometimes increasing operational expenses or inventory is worthwhile if it helps you increase throughput enough (and the increase is stable enough). Increasing them has an upside and you have to look at the cost and benefit.
This contrasts with, say, increasing reason and decreasing superstition among scientists. There’s (to a reasonable approximation) no upside to a few more scientists believing in ghosts.