Absent some reason to doubt the reported wind (as noted above), using a number that's 20kts off is no longer conservative, it's wrong.
Correct me if I'm wrong, but don't the field performance numbers assume that the actual headwind is only half of what's reported, and the actual tailwind is twice what's reported? I could be thinking of something else though.