Since I wrote about the mental goal system and the ninja/unicorn experiment, I've been thinking about more implications of this. Peter called it "the most difficult trap I think our civilization lays for us."
There is also a category of "unresolvable" goals, meaning that it is impossible to know whether or not you have really accomplished the goal. For example, as I commute to work in the morning I might have a goal of taking the quickest route to work. The problem is that I can't really know whether I succeeded. Even if I arrive at work in a reasonable amount of time, some other route might have been faster, if only I had taken it. So I will probably never get to enjoy the mental payoff of knowing that I accomplished the goal.
Even worse, though I can't prove I succeeded, it is very easy to suspect that I failed. If there is even a minor delay on my chosen route, I will tend to assume that I made the wrong choice, and a different route would have been better. Again, I can't prove this because I didn't take the other routes, and they may not have been any faster. But because the suspicion of having chosen wrong comes so easily, it is very likely that I will end up starting the day with a feeling of failure, however slight.
The flaw or trap is in the goal itself. Its very structure guarantees that having that goal is much more likely to make me unhappy than happy. Once I have that goal, there is nothing that can happen in the real world that can fix it. It is sort of like the "unicorn" goal but more subtle.
A more useful goal would be something like "travel to work safely and comfortably." It is possible to achieve that, and also possible to know for sure that I did achieve it. Having that goal is also more likely to make me happy than unhappy.
How we set goals makes a difference, especially considering that we can feel bad even about failing at an unconscious goal. I can't prove it, but I suspect that unhappy people probably have too many "unicorn goals" and happy people may have plenty of "ninja goals."
I also think that the Getting Things Done methodology is effective partly because it encourages a focus on next actions, which are straightforward goals that can be accomplished, and whose success state can be known. A next action is definitely not a unicorn goal, and having a large enough list of next action goals may prevent unicorn goals, simply because there isn't enough time left for them.
Anyway, back to my main point: unresolvable goals like the "fastest route to work" goal are harmful, and there is no real-world solution to them. Getting a faster car won't help. The only solution is to avoid having such goals, and to think carefully about the underlying structure of the goals you have.