Multi-objective optimization theory indicates that decomposing a complex objective into multiple sub-objectives and optimizing them separately is generally more effective than directly optimizing a single mixed objective.