@satrapper, I just tried scenario 2 of your post, and it seemed to work correctly. ie, I could get autoplay win single player, but not in multiplayer. So maybe the problem is a bit more subtle. Maybe it behaves differently depending on whether the second game is loaded from a save file or started fresh; I'm not sure. I'll test it again a bit later.
@Elkad
I think you're right about the counter-espionage thing. It currently is unfair, and it should be fixed. It might be a little bit tricky, because the counter-espionage counter is associated with 'teams' rather than 'players'; but turn order associated with players rather than teams. -- But I do think it needs to be fixed. I suspect there might be a couple of other things with similar problems.
I just had a look at when the city revolt timer is updated. Apparently it is updated at the start of the city owner's turn - which is good, but it still leads to some subtle & minor imbalances. For example, consider this: when a city is in revolt, the defenses are reduced to zero. So, if I cause a revolt during my turn, an I can attack the city in my turn and the city will have zero defenses. Now, here comes the imbalance: if the city's owner is the next player, then the city will be out of revolt immediately after my turn ends, and so no other players can make use of the reduced defense - but if the city's owner is the player before me, then the city's defense will be zero during the turns of the other players as well. Therefore the city revolt is more powerful when used against players before me compared to when used against players after me.
To make it more fair, I suppose the revolt timer would have to be associated with the espionage player rather than with the city's owner. But that would be a bit of a mess to implement, and it would also make the revolt mission a bit more powerful, because it would mean the city would be effectively out of action for longer - from the point of view of the city's owner.
That would be a tricky problem to fix... I think I'll just leave it alone for the time being, but I'll try to fix the counter-espionage one.
As for trade routes, I haven't changed the code for that and I haven't heard of the unfairness problem you've described. I have looked at the trade code a little bit while working on other stuff, and I vaguely remember that it is completely one-way. ie. The trade routes of each civ are completely independent of the trade routes of every other civ. Just becase city A has a route from city B doesn't mean city B has a route from city A. If I'm right about that, then I'm pretty sure the problem you mentioned doesn't exist. But I'll check it again to make sure, sometime.
I haven't looked into the pause thing yet.
The turn limit for the force civics espionage mission would be possible, but a bit cumbersome to implement... Maybe there's a better way to address the problem. (Presumably, what you're saying is that you think the force civics espionage can be a bit too powerful if used repeatedly. If the target keeps trying to switch back, then it costs them a lot more than it costs the espionage team...)