Lucky for you, this set is empty.
> ... tend to be less effective than...
At what? If you're an AI, then you might be overfitting on your training data.
Snide jabs aside. I usually find that "good enough, just execute" has its own failure modes, different than those of "keep digging, understand deeply". Better yet, there are no "failure modes" or "successes" at all unless you have a goal to evaluate against.