I don't think you understand the butterfly effect. The USA should not exist 999 times out of 1000 diverging from the year 1444 A.D. Neither should Rev. France or many other events. These are only given a slide since only few people grasp the concept of true alternative plausible history.
To draw a modern analogy, if Al Gore got elected as American President in 2000 instead of Bush, there would probably be no European Union, Ebola outbreak, iPhone, or many other modern topical entities. What is extrapolated from there draws infinite outcomes. Even if you decide to place your left foot even a fraction of a degree away in another direction, it causes a chain of reaction that leads to infinite change and outcomes over time. This change also melds with all the innumerable actions happening in the universe at any time.
"Chaos theory" and "period flavor" are nigh incompatible concepts in a GSG. Like it or not, people would prefer to fight against France with France-like borders at the end of the game with their westernized Malian colonial empire. DHEs, national ideas, and missions exist to nudge the timeline 'back on track', as it were, but player intervention is more than capable of changing anything and everything.
The ideal grand strategy historical game is one that gets a historical result the majority of the time without player intervention, and can spiral off in wild directions with player intervention. Whether you believe the historical result was 'likely' or not is irrelevant, since the simulation should, without any external variables introduced, play out mostly like OTL history. EU IV is really close to that.