"Thorough testing requires replication to verify results"? But if my data is "this burial shroud can be carbon-dated to 1200 AD", why wouldn't I be able to replicate that result? It'd be pretty trivial; just pick another sample area and test that. You're confusing the event with the data. The event cannot be replicated, but the data that led us to the hypothesis about the event can be.
Are you kidding me? I already did:
Not only is this simply not true, it also (and here's a big, fat, useful hint for anyone curious about how true these claims are) runs directly contrary to the way science has been practiced for the past 200-odd years. Tell this to a geologist, a paleontologist, an evolutionary-developmental biologist, an anthropologist, a climatologist, a cosmologist, an astrophysicist, or almost any other scientist in any other discipline, and they'll laugh you out of the room! You will not find any serious scientists who agree with the statement made by these people, because, well, it's asinine. It's like saying "science can make no statement about human emotions" - not only can it, but it has been doing so for quite some time, and I'm sure quite a few researchers would be quite interested to know that what they're doing is not science.
The statements are just wrong. PZ Myers also took them on head-on:
(Emphasis mine.)
No, I'm sorry, his logic is just straight-up wrong. As Myers points out:
There's nothing to debunk there. Ask any real scientist what they think about the matter, and you'll hear "yeah, you can make repeatable observations about the past". Or, to put it another way:
If I show you this picture, and your life depended on it, could you tell me to any degree of certainty what kind of natural disaster is going on here? I think you could, just like I think most people could. You can do that despite the fact that we can't repeat the events that led to that photo. We can draw inferences from what is contained in the photo, as well as concordant other lines of evidence, to conclude that this was most likely a hurricane. Much in the same way forensic science can look at a bloodstained fingerprint on the wall and the murder weapon and conclude that the murderer is whoever has those fingerprints (the fact that it's a forensic science lab claiming that you cannot use science to investigate the past is doubly disturbing). The observation of the fingerprint, the observation of the damage from the hurricane... These are things that lay in the present that point to an event that happened in the past repeatably and consistently.