Richardson-Gaudin models: What can we learn from (breaking) symmetry?