We are in the business of discussing plausibilities, likelihoods, their degrees, and what is and what is not plausible and not likely. And we fill in the gaps with narrative license. We can reasonably postulate what would happen for a while after the POD. It gets more complicated later on, because you're opening up all realistic possibilities when you deal with the multiverse; all are valid. My point is simply that you cannot assume history will go roughly or exactly as it did around the change, especially reflecting on the fickleness of certain among those decisions in the OTL. This is really a long runaround to just say that.
Interesting post and one I largely agree with. A good example might be the Kennedy assassination. How many people lived who would have died in car accidents that weekend had they not been home watching television coverage? So much of what happens in life is random and most AH tends to ignore this by holding all other factors equal. This is a legitimate analytical and literary tool, but if one were running actual experiments on changing events, one would probably find some surprising effects from changing big events. We might find, for instance, that a 2-year-old child named Barack Obama was killed in a car accident on November 24, 1963.
And the larger the event, the larger the later effects. Suppose there was no attack on 9/11. It is very possible that the world today would look very different, as the number of lives directly affected was immense, not only from the actual attack, but from the countless disruptions in people's lives afterward as flights were diverted and people's lives were disrupted. Strange as it may seem to think of it this way, but the events of 9/11 not only took lives, but it saved some as well. The degree to which this is true is ultimately unknowable without running controlled time travel experiments, but these possibilities are interesting to think about.