EDIT: Most ecommerce startups would fall into your "proven business model" category, and yet they, like McDonalds before them, do extensive experimentation to find the path to rapid growth. Which is why Steve Blank underscores that the search is never finished.
If a company was formed to search for a scalable, repeatable business model, then he's still interested in it when they have found it. Then they're in the growth stage.
But if a company was formed, as most companies are, intending to use a proven business model, then it's not in his definition of a startup. No matter how fast it grows.
This would seem to exclude virtually all delivery service, cleaning service, car service, etc. companies, which are usually considered the prototypical SV startups.
If you were to start a Postmates competitor today, though, and were just doing what they did, then it wouldn't be a startup. At $250m ARR, their business model is already proven.