Text this: Widespread analytical pitfalls in empirical coexistence studies and a checklist for improving their statistical robustness