It seems that falsifying would provide an escape route. Unfortunately it doesn't, because you would first have to establish - without a doubt - the truth of the

It seems that falsifying would provide an escape route. Unfortunately it doesn't, because you would first have to establish - without a doubt - the truth of the contradicting statement. That works ok-ish for collapsing bridges, but it breaks down for less direct observations, e.g., when reading experimental data from a computer screen, collected to refute a Physics theory.