fix: dampen agreement factor by sample size in trend confidence to prevent low-evidence inflation

Agreement of 1-2 signals was inflating confidence to paper-eligible
levels (0.575) even with low credibility sources. Added log2-based
dampener that scales agreement contribution by unique source count,
saturating at n=7. Single signals now cap at 0.39 confidence,
2 signals at 0.49 — both correctly below paper threshold (0.50).

tests/test_paper_trading_simulation.py

@@ -426,11 +426,11 @@ class TestRecommendationDrivenOrders:
         impacts = [
             ImpactRow(
                 document_id="doc-weak-1",
-                confidence=0.20,
-                novelty_score=0.1,
-                source_credibility=0.2,
+                confidence=0.40,
+                novelty_score=0.3,
+                source_credibility=0.5,
                 sentiment="positive",
-                impact_score=0.1,
+                impact_score=0.3,
                 catalyst_type="other",
                 key_facts=["Minor update"],
                 risks=[],
@@ -438,11 +438,11 @@ class TestRecommendationDrivenOrders:
             ),
             ImpactRow(
                 document_id="doc-weak-2",
-                confidence=0.15,
-                novelty_score=0.1,
-                source_credibility=0.2,
-                sentiment="negative",
-                impact_score=0.1,
+                confidence=0.35,
+                novelty_score=0.2,
+                source_credibility=0.4,
+                sentiment="positive",
+                impact_score=0.25,
                 catalyst_type="other",
                 key_facts=["Routine filing"],
                 risks=[],