> SpaceX had only a very small (single digits) number of test launches they could do before running out of money. A couple more grid fin unknowns causing failur

> SpaceX had only a very small (single digits) number of test launches they could do before running out of money. A couple more grid fin unknowns causing failures and SpaceX would have gone bankrupt so it wasn't clear back then. This is in no way true. When SpaceX was working on F9 reuse, their business was (very) profitable at their launch prices. They are still pricing F9 launches at a point where the launches would be profitable if every launch was expendable. Reusability was not what they needed to win -- it was the victory lap after they had already won. There is a lot of fud arguing otherwise based on external estimates, but they all fail to take into account that the internal cost of F9 is only less than a fifth of what all the competitors pay for their comparable rockets. The reason is that SpaceX is extremely vertically integrated, and every component that goes into their rockets is heavily optimized for cost first and performance second. For example, most other American launchers use RL-10's for their upper stage engines. There is a good reason for this -- RL-10 with it's best-in-class Isp is quite possibly the best upper stage engine that exists. However, this comes at a rather steep cost that is not public knowledge but was in 2017 estimated to be ~$25M. The costs of the rocket engines on a F9 is not public knowledge either, but is known to be less than $500k each, or at most ~$5M total.