But if what I am now thinking you are saying is correct I agree with you. I suggested it a while back but it would require a complete overhaul of the intel engine.
My idea was that instead of intel projects always suceeding unless there was suficent CI in which case they would always fail, that intel projects have a chance of sucess that was dependant on the operation, and the level of CI. More damaging projects would be harder to do (less chance of sucess), and take a minimum amount of time to acomplish. Throwing more points at the project wouldn't make it happen faster, as is the current system, but would increase the chance of sucsess up to a point. CI on the other hand would not guarantee failure of a project once a threshold was reached, but would decrease the chance of sucess depending on the amount of points spent on it. But every project no matter how many points you spend would have some chance of failing regardless of whether or not your enemy even had CI. And no matter how much CI you had the enemy would have a chance of "getting lucky" and sneaking one through.
Also with this system there could be "partial failures". Since a project has a fixed minimum amount of time to accomplish, a partial failure could set them back a few months and would reflect the agents having to go deep undercover or hide in safe houses. While a complete failure would remove all progress on the project and would reflect the operatives being captured, etc.
Also, I didn't think of this before, but since the projects have a fixed minimum amount of time to perform, these projects in progress should require a certain minimum maintenence in intel points, or they would go away, losing all progress up to that point. You could "put the operation on hold" so to speak, but you would still have to spend points to keep it avaialable, and the longer it's out the the more chance of being detected and thwarted.
Another thing I just thought of you could do with this system would be to turn the agents, as the British did in WWII. Keep them in place, but feeding misinformation to the enemy. The enemy would be spending point on the operation, thinking it was still working, but the points are being wasted because the agents are in control of the target empire. You could even give false Messages that facilities were blown up, or incorrect fleet information. That sort of thing.
Of course this is all a pipe dream. This would almost be a game in and of itself. "Intel Empires"
Geo