Text this: LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation Systems