Can complexity science support the engineering of critical network infrastructures?