[Slick]

I think that's a lot better; at least it's clearer. It is true that points will be wasted on the invalid project. To me, a player should use the event log and take appropriate actions for each entry. If he knows he has "repeat projects" selected, he should manually remove the project after he reads the log entry about the successful intel project. If he didn't remove the project once it succeeded, the next turn he will get a "failed" log entry for that project and that should key the player to remove the successful project from the queue. As far as the AI selecting a new valid target, that's probably better than wasting points on a dead project, but the AI is not known for his wise choices...