By: colin

An observation:
1 in 93 is a very misleading approximation of 1.1% because
1 in 91 is the closest approximation in the "1in X" format. (X is integer). In fact X can be any number from 87 to 95 inclusive. All nine of these will round to 1.1% (1dp) but 1 in 91 is 20 times more accurate than 1 in 93.
Admittedly this error is completely trivial in the sense that it _should_ not have any direct effect on any arguments above. So you might think I am a pedant. But my point is that once you see this error, and the fact that no-one has flagged it, it inevitably does diminish ones confidence in the quality of the rest of the discussion.