Even theoretically, you would then need to have perfect information every single other factor influencing relevant-demographics death rates, assuming you somehow magically know the exact relevant demographics. If there is even one other factor that is uncertain, you end up having to increase your approximation's margin of error proportionally to the uncertainty, and each missing data point is another power factor of increase in the margin. Eventually, it's much smarter to realize that you don't have a clue.

Now, take into account that you don't even know al... (read more)