Assessment of LLM Responses to End-user Security Questions