SPIKE-4: Document what ISO fields MDTranslator preserves and drops

**What:** The ISO-to-DCAT-US crosswalk embedded in MDTranslator determines which ISO 19115 fields survive the conversion. This crosswalk may not have been formally documented as a decision record for Data.gov. It is not clear which fields are intentionally dropped vs. accidentally missed, and agencies and downstream consumers (including GeoPlatform) have no reference for what to expect.

Using the Census concity record as one of many test cases, map the full MDTranslator output against the source ISO XML to produce a field-by-field accounting.

**Investigate and document:**
- Run a test case concity record through MDTranslator and capture the full output JSON
- For each ISO field in the source record, note whether it appears in the output, is absent, or is transformed/renamed
- Pay particular attention to:
- CRS (`MD_ReferenceSystem` / `referenceSystemInfo`)
- Spatial representation type (`MD_SpatialRepresentationTypeCode`)
- Scale denominator (`MD_Resolution` / `equivalentScale`)
- Topic category (`MD_TopicCategoryCode`)
- Data quality reports (`DQ_DataQuality` / report elements)
- Lineage processing steps (`LI_Lineage` / `processStep`)
- Feature catalog link (`MD_FeatureCatalogueDescription`)
- Purpose (`gmd:purpose`)
- Status (`MD_ProgressCode`)
- Categorize each dropped field as:
- **(A)** Intentionally out of scope -- no DCAT-US equivalent
- **(B)** Possible mapping exists in DCAT-US v3.0 -- not yet implemented
- **(C)** MDTranslator gap -- DCAT-US v1.1 field exists but MDTranslator does not output it

**Done When:**
- Field mapping table produced and attached to this ticket
- Each dropped field categorized A, B, or C
- Category B and C fields become candidates for IMPL-6 scope