The LLM will spew out an answer in a format that most closely matches the average answer to your query in its training data. Most explanations of grammar look like your example.
You may be able to simplify the output with some careful prompts but ultimately the LLM has no understanding of grammar itself so it can't help you understand it if it doesn't contain enough training data that meets your requirements. Just like it can't answer numeracy questions perfectly.
Not at all.
The LLM will spew out an answer in a format that most closely matches the average answer to your query in its training data. Most explanations of grammar look like your example.
You may be able to simplify the output with some careful prompts but ultimately the LLM has no understanding of grammar itself so it can't help you understand it if it doesn't contain enough training data that meets your requirements. Just like it can't answer numeracy questions perfectly.