In The Art Of Game Design, Jesse Schell (an experienced game designer) lays it out for you, and if you look at the chapter list you can see right there in Chapter 2: "The designer creates an experience."

I snubbed this section because it sounded far too abstract. I wanted to get down to the nuts and bolts of building. However, having come back around again, I have to say that ignoring this fundamental undermined my ability to come up with a coherant design spec. Until I had the experience squared away, I had no idea what I was designing, and consequently I kept coming back to my factions wondering what exactly I was going to make them do and kept changing my mind because I had yet to really decide on the experience.

So I've come up with the experiences I'm doing, and basically I'm actually developing three different experiences within the same game that vary depending on what you play:

Colonists/Terrans - Real time strategy style building game.
Aliens/Natives - "Sim Ant/Dwarf Fortress" style building game.
Privateers - Vehicular combat RPG.

You might say at this point, "Good God, Geldonyetich, why would you task yourself with creating three games at once that interact with eachother?" And you'd be right: biting off more than you can chew is a death knell for most indy projects, and trying to stick three separate types of games in a blender and make them work together is exponentially nastier.

Well, the reason is threefold:

1. First, because I won't settle for anything less. I've been gaming for 26+ years and I'm picky as holy hell. Another of Schell's chapters is titled, "A designer has a motivation" and my motivation is I want to make unique, epic games because I'm tired of all the cloned crap out there. After all, this is "My Own Net Dream" we're talking about, it's fair enough to suggest that if it's not something I'd want to play it's not something I'd want to make.
2. Second, because that's actually sort of how the game shook out because of all this code I've got laying around. A lot of this is already done, so I might as well just stitch it all together and call it done. It's mostly GUI work and a few holes to fill at this point.
3. Third, I think it's pretty nifty I've actually have figured out how to make all three of these games work together in pretty good harmony, so it'd be a shame not to shoot for it.

So, the experience I'm shooting for figured out, I expect the rest will fall right into place. The biggest hurdle I can see (outside of the usual bug quashing and balancing) would be a save game system, and that might not be too much of a hurdle considering I'm already using Swapmaps.