fix: regenerate golden hex for dual-16pt Doppler and add real-data TBs to regression

Regenerate all real-data golden reference hex files against the current
dual 16-point FFT Doppler architecture (staggered-PRI sub-frames).
The old hex files were generated against the previous 32-point single-FFT
architecture and caused 2048/2048 mismatches in both strict real-data TBs.

Changes:
- Regenerate doppler_ref_i/q.hex, fullchain_doppler_ref_i/q.hex, and all
  downstream golden files (MTI, DC notch, CFAR) via golden_reference.py
- Add tb_doppler_realdata (exact-match, ADI CN0566 data) to regression
- Add tb_fullchain_realdata (exact-match, decim->Doppler chain) to regression
- Both TBs now pass: 2048/2048 bins exact match, MAX_ERROR=0
- Update CI comment: 23 -> 25 testbenches
- Fill in STALE_NOTICE.md with regeneration instructions

Regression: 25/25 pass, 0 fail, 0 skip. ruff check: 0 errors.

9_Firmware/9_2_FPGA/run_regression.sh

@@ -403,6 +403,18 @@ run_test "DDC Chain (NCO→CIC→FIR)" \
     tb/tb_ddc_cosim.v ddc_400m.v nco_400m_enhanced.v \
     cic_decimator_4x_enhanced.v fir_lowpass.v cdc_modules.v
 
+# Real-data co-simulation: committed golden hex vs RTL (exact match required).
+# These catch architecture mismatches (e.g. 32-pt → dual 16-pt Doppler FFT)
+# that self-blessing golden-generate/compare tests cannot detect.
+run_test "Doppler Real-Data (ADI CN0566, exact match)" \
+    tb/tb_doppler_realdata.vvp \
+    tb/tb_doppler_realdata.v doppler_processor.v xfft_16.v fft_engine.v
+
+run_test "Full-Chain Real-Data (decim→Doppler, exact match)" \
+    tb/tb_fullchain_realdata.vvp \
+    tb/tb_fullchain_realdata.v range_bin_decimator.v \
+    doppler_processor.v xfft_16.v fft_engine.v
+
 if [[ "$QUICK" -eq 0 ]]; then
     # Golden generate
     run_test "Receiver (golden generate)" \