What would it take? How do you know if the mechs have not taken over the world and stuffed you into a pod where all your senses are controlled by the matrix? How do you know you are not an avatar in an alien video game where he who dies with the most yields the most points? How do you know you were not created last Thursday with memories intact? How do you know God did not create dinosaur fossils as a test? At some point, you have to go with what is reasonable.
Design proponents claim that we simply do not know the function of all the DNA which may not be directly involved in coding or identified regulation. Here you suggest the GULO deletion segment could fulfill a purposeful design. The problem with that idea is that this is not a stretch of DNA for which we do not know the function. The question posed by the GULO deletion is not just, “this seems to have no function so why would anyone design this?” We do know what the purpose of this DNA is. There is not some mystery or ambivalence. We know what it is supposed to do, we know how it is broken, and that it is useless for anything else. The “maybe it is a purposeful design” counter argument doesn’t make sense, because it does have a purpose and to that purpose it is non-functional.
The odds of getting a particular losing lottery number are the same as the one that wins. How many frameshift mutations have happened, but not been fixed? Perhaps millions? Some mutations must be fixed.
But you raise a good point. John Woodmorappe and Jeffrey P. Tomkins have reasoned that the shared mistake of the vitamin C deactivation is just a case of lightning striking twice, and that the ape and human mutations were independent. As you suggest, the odds are against such mutations being fixed. To have it happen independently, and in exactly the same way, are very much greater than a common instance and is really devastating to their argument.