I worked on the first demo of Star Citizen
a decade ago, prior to the announcement and up through the first Kickstarter. I haven't followed the development much in recent years. I feel like there's been a joke circulating for awhile about finally implementing a "roadmap to a roadmap". Here's my (very limited) insight into how the project took the direction it did.
Chris Roberts is a brilliant guy to work with, a first rate coder and by any standard a visionary. He seems to have a new idea every ten seconds, and they usually begin "it would be cool if..." An hour later, you're still taking notes and he's saying, "and it would be realllly cool if..." Technical difficulties and obstacles were, at that stage anyway, actually a reason to do things. Part of the idea was to prove that PC gaming would always outstrip the console experience. Polygon counts could be made higher, and higher again. The goal wasn't performance, but the opposite - how much could you put on the screen. Think about it now with 10 years of hindsight, that's a race you're only one ever one generation of consoles away from losing.
My impression of working with him was that there's no such thing as practical constraints. A good enough coder should be able to get around anything. Prior to the first demo, there was a hard deadline to have something to show, so a lot of corners had to be cut. Once that happened, his attitude immediately shifted to "sky's the limit". And with the wild success of the first round of crowd funding, that became "spare NO expense." This was never just a space game, or even only his magnum opus. It's a fully functioning universe that exists in his mind, that just needs to be turned into code. Everything Wing Commander was limited by, this would be unconstrained by. Everything is possible as long as you're committed to always building for the latest platform, on the most powerful machine in the known universe, and your audience was committed to having the same hardware. A newer engine comes out or newer hardware? Tear down everything and rebuild for it.
So, I never thought that this was a crowd-funded game about the making of a game. I think it's what happens when perfectionism meets unlimited resources. I also have to say that the community forum was very, very involved and demanding from the beginning. They were already playing the game they imagined in their own heads. Having put their dollars in, they felt entitled to their say in feature development. And this crowd involvement, their one-upping each other's ideas, mixed with Chris's almost infinite drive to incorporate new features, and his impulse to say yes to almost anything that sounded cool, and then a tsunami of cash, created a dynamic of setting up more and more far-fetched goals right out of the gate.
Around that point, what I had built - which was never actually integrated into a working build - was jettisoned for a different approach. He hired 500 people and I sort of drifted off without ever being specifically let go.
Still, it was a great experience, one of the most challenging and certainly one of the most fun projects I've ever worked on, especially in the sky's-the-limit phase, and Chris is the kind of demanding but fair person who held everyone to high standards, but who held himself to the highest. He was a pleasure to work with. I hope he finishes his game.