It's a critical distinction to draw: informing a backtest and demonstrating empirical concretization are two separate classes of outcomes. Your point on "preuve tangible" in France nails the demand for rigorous classification.