It's key, and it's clever design. Basically space shuttle next edition with the booster on the bottom rather than it being strapped to the side of it and where the booster is reusable. It's a two-part vehicle: booster/shuttle.
The original space shuttle had a 40% vehicular failure rate. The SpaceX Shuttle need to have commercial airline rates of failure and reusability. That's a big step up.
You know the way we still have 80 column terminals because punch cards had 80 units?
The decisions SpaceX make now are going to become space-faring standards for decades if not centuries to come.
To frame in a comment we often hear around here: if you don't control your funding, you don't control your destiny.
Building anything for the US government that's big enough and has potential military applications is virtually inviting them (and the large project procurement morass they bring) to become involved.
The Soviet Buran shuttle had the right idea. It didn't have main engines and was launched with the Energia rocket, which could be used independently with other cargo as well. Also, it was capable of fully automated landing, unlike the American Shuttle
From a commercial, ROI standpoint, having %40 of your hardware eventually destroy itself is bad news. Especially since each time your whole enterprise stops for years until you find out what went wrong and fix it.
The original space shuttle had a 40% vehicular failure rate. The SpaceX Shuttle need to have commercial airline rates of failure and reusability. That's a big step up.
You know the way we still have 80 column terminals because punch cards had 80 units?
The decisions SpaceX make now are going to become space-faring standards for decades if not centuries to come.