And yes the risks absolutely would "rise sharply" with distance, that is the nature of an exponent. Those ideas arent in question.
What is in question is that we underestimate this risk, or that it exceeds a linear estimation, which are not proper descriptions.