I think a great example of how those variables come into play is seen with ash (specifically green and white). When they are healthy they really only commonly fail at a co-dominate - and quite easily at that. Otherwise, they are very strong. When they are sawn and dried, they are one of the stronger woods being preferred for baseball bats and shovel handles. However, when EAB kills them there is something else at play (I strongly suspect a decay fungus, but haven't seen anything identified...doesn't mean it hasn't been, just not last time I spent the energy to look!). Ambrosia beetle gets in there, so they may be vectoring something else? But anybody working with EAB killed trees knows they are not stable. Loggers won't take them because they split as soon as they go through the mill.
So the "ideal" Species failure profile for Fraxinus americana and F. pennsylvanica would discuss their strength when healthy. Their slightly higher than average likelihood of splitting at codominate branches, the probability that they will be killed by EAB and secondarily infested by ambrosia (not necessarily unique...), at which point they become excessively brittle (caused by ???) and may just as likely break off at the root ball or 15' up, but won't likely stand as a sturdy trunk for many years slowly shedding branches.