Benchmarking LLM-Assisted Blue Teaming via Standardized Threat Hunting