docs: frank-large FA_TQ A/B rejected at 80k; deep roles rerouted to frank-primary

Stop-window A/B, same 80k filler prompt, 128-token essays, runtime
timings: HALOGEN_FA_TQ=256 (upstream issue #12's carve workaround) moves
decode ±0.3% (noise, draft acceptance identical) and costs 9.5% prefill
on this topology. The depth collapse sits deeper than the FA tile knob;
the default stands. The harness's slow/advisor roles moved to
frank-primary (fp8 KV 7.32 KiB/token = 8.7x less KV read at depth).

serverAppConfigs-docs/LLM/index.md

@@ -89,6 +89,11 @@ frank-small, port 7771]
   run`; the `-d` era wrote only the container ID). Decode-at-depth is
   structural: 3.3-3.8 t/s at 80k vs 30-50 shallow; no supported flag
   touches it.
+  The FA tile lever from upstream issue #12 (`HALOGEN_FA_TQ=256`) was
+  A/B'd at 80k on 2026-10-06: decode ±0.3 % (rejected), prefill −9.5 %
+  on this topology — the default tile stands. The deep harness roles
+  (`slow`, `advisor`) rerouted to frank-primary; frank-large keeps
+  OpenWebUI and shallow callers.
 - [[frank-small-llama-swift15-iq3s-mtp-131072]] — **ACTIVE since 2026-10-05
   (evening); 2 × 131,072 no-spec since 2026-10-06** (root systemd unit,
   port 7771): llama.cpp master (CUDA sm_86) serving ukisai/Swift-1.5-