Mr. Collins cherry-picks examples from very disparate enterprises and markets to attempt to arrive at a general set of market-spanning, culture-spanning, and era-spanning operational principles. It's bad advice and worse--bad data science.
In my personal business experience there were also several instances of "one killer innovation" or "one lucky break" that, we realized both at the time and in retrospect, really did "make" the business.