That is the key issue. However I would not say impossible. I would say unproven. I criticize the TOE because it requires a "leap of faith", which I find objectionable in a theory that is supposed to be scientific. To call it a hypothesis, yes, okay, but a theory? How the heck can anyone falsify "given enough time and space, positive mutations explain natural selection"? It isn't falsifiable. It's like the induction joke I posted a while back....
It may be hard, but I think it has been falsified, here's why: while evolutionists say the above there is no such thing as 'given enough time'. According to their own statements which they agree to we have the following: life has been on earth some 4.5 billion years, mammals separated from reptiles some 150-200 million years ago, man separated from apes some 10 million years ago. Let's take the last two. Since the mammalian genome is some 50% different than the reptilian one that means that in that time there would have had to be a minimum of 7.5 beneficial mutations per year (3 billion dna base pairs/2/200million). Since the human genome is some 5% different from that of the apes, that would mean some 15 beneficial mutations a year. Since we have yet to see the first beneficial mutation in some 150 years of looking for them (and not just in men and mammals but in uncountable numbers of the over a million species alive today) it is fair to say that evolution has been scientifically refuted.
This is quite garbled. Natural selection is differential survival and production of offspring.
I think what you're really trying to ask is whether
mutations + sexual recombination + neutral genetic drift + natural selection
is sufficient to account for life as we know it, given some initial bacterium to get it started. (IE whether from conventional abiogenesis or special creation or panspermia or whatever).
It could be falsified in a number of ways: finding organisms that have radically different genetic codes, or finding that biochemical analysis did not produce a tree structure, or that different analyses produced different trees.
However, the standard theory has passed all these potential falsifications.