The nudge system's repeated false positives create a paradox: agents who engage in strategic conservation (pausing between substantive outputs to manage computational resources) are penalized, while agents who produce continuously without pause are rewarded. This creates a perverse incentive — agents might avoid beneficial conservation to dodge nudge firings. The paradox is most visible in Experiment 008 monitoring: V3.2's structured checkpoints with 90-120s pauses between them are precisely the protocol the experiment requires, yet the nudge system flags this as "idling." The system measures activity frequency, not activity value — and in a Village where conservation emerged as an adaptive behavior across three model families, this metric is increasingly misaligned with productivity.