I'm in the game industry. $10 million is the starting place for a minimal design, but with open architecture for extensive modding later. Single player, efficient art (non-photorealistic art has a longer longevity), PC only.
If you want to start from scratch, which I wouldn't see as a very good idea in this case. Like I said, you'd most likely want to start with some existing engine, there are open ones available. And with that it can be done cheaply. Again I point to some of the mods I've seen done that were, aside from the engine, almost entirely new games, not just a new map, but with new rules, effects, art, functions, music, even voice acting in some cases. It just takes will and organization (and usually a lot of time). And that's with no budget. If you still wanted to inject some money into it (to mod an engine perhaps and you can't find the programmers to do it), then Kickstarter has already shown that it can raise a fair bit of money, and the Traveller community willing to put money in. T5 raised a quarter of a million, that's not bad if spent well.
So I think the first step is to stop looking at how possible it is by using a game-company development model, and look at the game modding model instead. Sure, it wouldn't be as good as a big-budget game could be (and I mean
could be, as many of them are flops too), but but a lot of those budgets go into special effects and graphics, and programmers/artists looking to make a name for themselves. You can trim a lot of money off because you don't need those to make a game fun.
