Understanding the Synergies between Quality-Diversity and Deep Reinforcement Learning