It was worth it in the case of Airbnb, Twitter, Teespring, Coinbase and many, many others. Once you find product market fit, you can afford to rewrite your entire backend if needed. If you don't win round one, then it doesn't matter how well your stack could have scaled.
The time you save initially you'll later spend fighting the framework or debugging its internals.
Python's "explicit over implicit" scales a lot better, especially with large teams.
(personal opionion of someone maintaining production-critical Django and Rails applications)