That's fine but as I see it, what you have outlined only looks at various factors in isolation of others.
Accidents tend to happen not just as a single failure in design or decision, but because two or more things have gone wrong in a sequence of events.
So I am talking about risk assessment where the combinations of known limitations and factors (that in themselves may be acceptable) are examined to see whether there is any factor that, when interacting with one or more other factors, takes the overall risk factor above an acceptable level.
Of course there will always be the 'unknown' risks, but if you have a sufficient margin of comfort with all the others, then you are better able to deal with a sudden emergency than when you are right up against the limit of safety and then something unforeseen occurs ...
[ 08 September 2001: Message edited by: Seat 32F ]