Amazon can fail a lot because money for private companies is a lot more expensive and harder to get than for public companies. Musk is playing on hard mode because he wants to retain control.
Musk claims he won't take SpaceX public because public investors have short attention spans and would require profit/return too soon. Because he wants goals set out further than the next quarter, he has kept the company in the hands of like minded investors.
That may just be bloviating on his part, but it does seem at least sensible on the surface.
It means the shareholders decide who is control. This is not about having control now, it's about retaining control.
Musk doesn't want to be beholden to shareholders because their interests (profit) might not align with his objectives (Mars). Presumably Bezos does not fear any such misalignment of interests between himself and his shareholders.
You need both. A visionary alone needs the discipline that comes from having to choose which vision of many and finish it. A financial focused manager will never bet on any but the most obvious ideas and thus lose when someone else makes something work.
Ideas are easy. Implementation is hard. The least creative person has enough ideas to bankrupt even the biggest company if you start a team to implement them all as the idea occurs (a car obviously takes more people than a paper notepad, but you should expect to have all those ideas)
A visionary who never thinks about money will eventually run out, or worse bounce between ideas and never finish one. The person who only thinks about money will not invest in the things required to continue making money.
I've heard it stated that you need a visionary CEO and a close second who is a financial realist. The visionary has lots of great ideas, the financial realist reigns them in and forces the visionary to choose which vision(s) to focus on and finish; and let the others go. This works great until the visionary CEO steps down (dies, retires...), when you put the realist second in charge all the visionary projects are stopped, for a couple quarters the company makes more money, but there are no visionary things in the pipe so the company dies a slow death.