Is that a fallacy with basic algorithm questions, or with interviewing in general? How does another category of question provide objective feedback about problem solving ability instead?