Add agent rehearsal harness, findings, and per-run evidence for split-loop-with-lambda

Six scripted pi sub-agent runs (3 tool regimes × 2 models) rehearsed the
exercise; deterministic gate = solution test + approved file. Key results:
mandated CLI triples cost on the strong model (4:06 vs 0:39, PASS both);
weak model fails only under the mandate (A+Q), passes when free or optional.
Drift documented with code: scorer.updateSelection left inside the extracted
method (A+Q, C+Q). Harness in tools/rehearse; summary in
rehearsal-findings.md; per-run evidence in rehearsal-runs/; cheat sheet in
rehearse-with-agent.md corrected (lambda-typed parameters unsupported).

ideas/split-loop-with-lambda/exercise/rehearsal-runs/B+Q/verify.txt

@@ -0,0 +1,10 @@
+--- Run B+Q ---
+final diff vs pristine: 39 changed lines -> /tmp/rehearsal/runs/B+Q/diff.patch
+mvn test with solution's test + approved file:
+  RESULT: PASS (deterministic)
+structural checks:
+  [x] method collectExpectedWeather exists
+  [x] takes LongFunction<Boolean> parameter
+  [i] note: 'scorer' still referenced in LongRangeForecasts.java
+  [x] lambda syntax present
+  [i] main class differs from reference solution (may still be a valid variant)