Evaluating the Usability of LLMs in Threat Intelligence Enrichment