I think it's a very useful distinction. I've seen similar things in demographic statistics. Where e.g. parents education was more predictive of their child's outcome than their income. Career choice and success is somewhat noisy and random.