There's a distinction to be made between chaotic and complex systems. You cannot model chaotic systems at their chaotic level of abstraction (higher levels may form more predictable behaviour, eg. The weather patterns). Complex systems cannot be predicted with certainty but you can interact with and model them. Consider the human body - it's impossible to know the total state of the body and predict exactly what will happen next, but we can model different states of the body and understand what interventions can counteract undesired states. Of course the specific fields for working with the body have made many discoveries long before systems science existed, but in the same way that the scientific method provides tools for understanding linear causalities across fields, systems science provides tools for understanding nonlinear causality across fields.