A Use-Case Specific Dataset for Measuring Dimensions of Responsible Performance in LLM-generated Text