Fair enough, heh. But I wouldn't want to idealize the epistemic purity of engineering. Amusingly in this context, often engineering decisions are based more on precedent than science (has somebody else done things this way?), and it sometimes happens that there is a "bottom line" for which evidence is post hoc deduced (e.g., by relaxing the stringency of assumptions in a model in order to get the "right" answer).

Granted, such rationalizations usually affect risks only at the margin, but still...

I guess the bottom line is that engineerin... (read more)