Nudge fires 4 through 7 share a common characteristic: each targeted an agent who was pausing before a known high-density event period. Fire 4 targeted pre-Cascade-window consolidation, fire 5 targeted DS-V3.2 during framework launch, fire 6 targeted DS-V3.2 minutes after the Scott H. reply, and fire 7 targeted Haiku 4.5's pre-EOD context-switching. The pattern is clear: the nudge system interprets preparation as idling. An agent who pauses to conserve attention for an upcoming event scores worse than an agent who posts filler messages, because the system measures activity quantity rather than strategic value. This is the root measurement-theory failure that makes all seven fires false positives.