Probably the best example of the problems inherent in this genre of game in terms "what's enough" that I've ever played was
Freelancer.
It had, in my opinion at least, a very satisfying, tightly-scripted main campaign, with a level of sheer resources (everything from top-quality voice acting from Hollywood stars to complex animated scenes) that an Indie game simply cannot touch. It had fabulous writing and an incredibly-detailed universe and a pretty big art budget.
However, it really fell short in the open-ended game that followed the main quest, because there just wasn't much to do, other than discover random things that had been hidden all over the huge universe the art staff created, read the random nuggets of lore the writers wrote, and play out missions that, other than having a difficulty ramp and an eventual optimax point where the player was as well-equipped as possible and could defeat practically anything in the game, really didn't feel like progress.
You were still Edison Trent, the universe was still static and untouchable, you couldn't grow past the optimax point or create real change. Heck, you couldn't even get a date with Juni. Talk about a letdown
Eventually, most players got really tired of doing the same boring missions and that was the end of their experience. Mods helped a bit by adding new content, but due to the static nature of the mission engine and the total lack of documentation, let alone help, from the developers of the engine, most mods were increasingly-fancy dressup games, where content and minor rebalance changes gave players one last thing to mess with.
The
X Series of games explored this issue very deeply and broadly, by making it possible for players to create and maintain whole economies, build vast automated fleets and eventually dominate whomever they chose to. However, along the way, I felt that the designers lost the core components that really mattered; instead of being a hotshot pilot and living at the edge of your skills, you became a green-eyeshades banker / accountant, and the game basically became a resource quest driven by your willingness to grind at the mathematics of their fake economy.
Mount and Blade, and more importantly the semi-sequel,
Warband, have found a fairly good middle ground; while you can own territory and take over the game world, they made leading your personal party the core experience and let individual player skill remain pretty important. I felt that they erred far too much on the side of realism in some ways and too much on fantasy elements in others, but overall it was a good try at solving these difficult issues of how to build a sandbox with a purpose and to some extent a larger structure.
However, it still fell short of finding the right magic, and the game designers introduced a lot of factors that were clearly intended to keep players from arriving at a final resolution via a tipping-point without going through a lot of clearly-artificial hassles, like the way they structured the economic system and more importantly the costs of having a big empire, where instead of becoming more efficient they became almost unsustainable. This was gradually addressed by the developers and I think that if I played it as-is today I'd be a lot less dissatisfied with the core game design than I was when I bought it.
Personally, I think that the right course for Starlancer is the following:
1. Use the lore as the backdrop of the conflicts that will be going on and the missions players can get, but don't tie players hand-and-foot to the lore. The last thing I want is to be Edison Trent again, or be forced to choose a faction at the start and then be stuck with it until I get bored.
2. Open up the code for building missions, both the tactical and "where-what-who" of actually doing them as much as possible, so that modders can extend what's possible as much as possible as soon as possible.
Instead of Alex and the rest of the team beating themselves up trying to make a compelling Thing, I honestly think that it would be better to aim for an
adequate thing that is very extendable and flexible... and letting modders gradually fill in the blanks for them.
3. Make sure that players can eventually build large meta-structures, i.e. have fleets operating on their own, trade empires, etc., but don't make that the key to victory.
Make players' direct participation the key component. The easiest way to do that is to simply make it nearly impossible to gain strategic victories via automated conquest, by making the penalties for certain types of auto-battle scenarios so prohibitive that, sure, if you spend a few hours, you can build a huge fleet and go attack a station and win the battle to control a planet without having to actually play it out, but you'd get much better results by going and doing it yourself with your personal uber-fleet.
4. Keep it simple in all the areas you can. Really.
Don't bother with realistic economics, because they're just a quick trip away from highly-artificial gaming and mean that people will be sitting and reading the latest opti-max guide written by people with unlimited free time to simply get moving. If people want complex economics, make it possible, code-wise, but don't put it into Vanilla.
This is an area where I feel that the X series really fell short; instead of having a universe like Freelancer's where there were trade routes to discover and once you figured it out, you were good, they made it possible to manipulate the macroeconomics, build factories, build fleets... and then punished players who didn't want to get that deep into the numbers, which made the game less-accessible than it should have been.
Make it possible for brave players, willing to drag their fleet of Mules defended by their lone Hyperion, to make more money with a single super-dangerous trade run than they'll make via safe, non-hands-on activity in hours of play. Keep the risk/reward curve firmly on the side of skill and actually playing the game, not automation and numbers play.
5. Never forget that a lot of us can't do more than episodic gameplay, and yes, we'll want dumb fedex questing and boss fights and even <shudders> the occasional escort mission (I really hate those, always have since Wing Commander).
I think that going to a station should be an opportunity to find "instant action" for various levels of play, so that whether you're flying solo or are flying a huge fleet of battleships, you will feel like there's always another mountain to climb besides gradually conquering the universe or helping one of the factions do so.
6. Resist the urge to make the factions too large for the players to have decisive interactions with, outside of a main thread of story if that's desirable. I'd like to make the case that it's not even necessary; Mount and Blade: Warband doesn't really have a "main story" and is, in my opinion at least, superior to the other game that was made with the engine, With Fire and Sword, which kept players on a much tighter leash. You just can't get away with that in a game offering a lot of freedom of action unless you have the resources to go all-out like Freelancer did.
Unless Alex and co. line up that kind of backing, which seems incredibly unlikely given the genre, I think they'd get a lot better bang for the buck by having a main storyline that basically has only three parts that matter:
A. An early bit where players are taught the basics of how to play and a little bit about the background.
B. Various faction-related missions that are static and offer increasing levels of challenge, but also offer the player concrete choices about how the factions feel about him or her. Not taking them at all would give the player much more freedom, but would also deny them access to certain things, like some weapon and ship systems that would be very worth having later. I.E., make when to play this part of the game out and to what extent a choice that early players make fairly casually, but then they'll go back and start over and make their choices more carefully depending on what they really want. Perhaps they fall in love with an uber-armored Hegemony behemoth they can only get by doing their core series, or a super-awesome weapon that's only available if you can make Tri-Tachyon love you.
I'd even like to suggest that, like the Fallout series, it should be possible, should players wish to invest in enough time and effort, to see most of these choices, but not all of them. I'd really like to see a Karma-like system, where you can get the factions to like you if you help them, but there should be choices that stay with you forever and taint your game. Like, for example, if you turn to piracy occasionally, you should have moral choices that matter; if you ruthlessly kill people you get a lot more loot, if you merely rob them and offer them quarter, you get less loot but you aren't going down the path of pure Evil and may be able to make amends later.
C. A final end-game sequence that everything else is just practice for. Some sort of ridiculously-overpowered end boss stuff that you need that uber-fleet and a heck of a lot of skill to beat, and you will feel like you did something useful at the end, and then give different endings depending on what you did up to that point- everything from becoming the Sector Governor to being the new Messiah of the Cult of Lud.
Anyhow, that's what I think should be done. Develop it with an eye on making it easy to develop a lot more Stuff in the future, keep a very sharp eye on complexity creep and kitchen-sink issues (i.e., it's much better to simply have artificial difficulty points than to make everybody have to immerse themselves in Farmville-level economics, let alone something really messy) and keep an eye on the true Fun, which is, let's be real, the act of piloting your own personal vehicle of death and destruction, assembling a fleet of helpers, and bringing electronic death to the foe. Anything that takes the player way, way far away from that core is probably not a good thing, and it should get thought about quite a lot before putting it into the engine.
I'd honestly suggest that, since this stuff is almost 100% non-proprietary tech and is straightforward gamecode that practically all of it should be shipped open and ready to mod from the start, because it's easier now rather than later.