...
1.Stable multiplayer engine.
Because there are no single player games, right? Considering the low user base we have to deal with here, any Orbiter-based combat sim would have to focus primarily on the single-player experience.
2.As was already said, a damage, temperature and energy production (for energy weapons and sensors etc.) model.
Agreed.
3.Reliable collision detection.
This doesn't have to be done in a fancy generic "it hit this precise spot on the mesh" way. Many games have successfully used a hit-box concept, so you'd have a hit on, say, the left wing. All the models used in the game would then need to have such a structure. A full-on super-precise generic collision detection engine is overkill.
4.Scientifically grounded physical models of weaponry.
What do you mean by "scientifically grounded?" If you don't allow at least a little bit of wiggle room, the game will end up being boring.
5.Surface bases/ships/... that can provide fire support
Concur.
6.Sensors/countermeasures/detectibility models
There is no stealth in space, if you want to be realistic about it. Some form of sensors other than just the built-in name tagging of vessels would be needed.
7.Set of standarts for spacecraft characteristics balance, that will limit and define the "development" of new weapons. E.g. one pound of engine weight of a given technology level and propulsion principle may produce X newtons of thrust; one pound of weapon weight may give out Y joules of kinetic/light/... energy per second.
Definitely. And you need a way for "illegal" craft to not be used (considering how easy it is to develop addons for Orbiter, you'd need a way to make sure that someone didn't go and throw on whatever they wanted). This would not be that difficult, I think -- if a player fires on something, query the player's vessel. If it doesn't respond correctly, then ignore the shot.
I can also think of a major thing you're forgetting that would make or break any combat sim: a story. You have to have something believable that tells the player why they are fighting. The story has to advance over the course of the game. This really isn't a technical limitation, but the game world won't be able to be completely static, I think.
Also important would be good-looking terrain, at least in the vicinity of surface bases. See below.
And then of course there's the question of how realistic space combat on an interplanetary scale is. Space is
huge. In general, a battle in the traditional sense would only occur at predetermined sites, never in deep space. By "predetermined sites" I don't mean that the two forces sit down and decide where to fight, I mean points of strategic interest -- surface bases, key orbital platforms, etc. (That's where the terrain comes in--a lot of the fighting will have to be on the surface).
This isn't "Firefly" where a Reaver ship going the completely opposite direction drifts past at a few meters per second. In reality, two vessels going opposite directions would have huge differences in velocity. It wouldn't be economical to completely waste all your fuel in order to slow down and engage the enemy in a traditional tactical engagement. You'd take a couple of very carefully aimed pot-shots at each other as you passed and hope to cause enough damage to prevent the enemy ship from doing whatever it was heading to do.