Mirrors PR #20 into the `coupler-CapNInterdigitalTee-cap_matrix.json` dataset so the NCap and CapNInterdigitalTee tables stay synced (same physical capacitor, two different `coupler_type` labels downstream consumers depend on).

Contents identical to PR #20 except for the single-field swap `design.coupler_type: 'NCap' -> 'CapNInterdigitalTee'`. Same 346 new Ansys-LOM records, same `design_tool` normalization to 'qiskit-metal' across all 464 Sadman-uploaded entries (118 already in main + 346 new).

Validation summary (vs main):
- 894 records (548 existing + 346 new), all 548 existing CapN records preserved byte-for-byte modulo design_tool casing.
- design_tool : {'qiskit-metal': 894}
- coupler_type : {'CapNInterdigitalTee': 894}
- sim_options block byte-identical between existing and new.
- 0 negative / non-finite caps, 0 Cauchy-Schwarz violations.
- Monotonicity: 0 / 363 cap_gap->top_to_bottom violations, 0 / 1803 finger_count violations; 27 / 541 finger_length->top_to_bottom violations all at finger_count=1, max 1.99% (~25 mfF, below Ansys LOM noise floor).
- Noise-floor sigma test (robust sigma from LOO on existing 548): median |z| < 1, p90 |z| < 3 on every capacitance.
- 0 duplicates / near-duplicates in the dataset.

Recommend merging together with PR #20 to keep the two coupler tables in sync.