How can one isolate the fixed time of death from other events surrounding it?
So let us see now. The argument often given is that if a given person had not died because of abc cause they would have died of xyz cause. Fair enough this argument seems to work for individual cases. Especially when these incidences are distributed over different geographical areas and time-lines in a normalized pattern. But it doesn't work for large numbers in a very small geographical area in a given instant of time.
And that is a huge problem.