My initial impression is this - we know that some of these things are the product of intelligent agency. Specifically, 1, 2, 5, and 9 were produced by humans, and thus must be the product of intelligent agency. But the problem is, the answers given in those cases were inconsistent - you determined 1 was designed, 2 was not, 5 was, and 9 wasn't. Or, more accurately, in the cases of 2 and 9, you were unable to conclude that they were designed artifacts, when we know for a fact that they are.
This does not really inspire confidence in the design inference, as I sure you can imagine. After all, if it has what appears to be a 50-50 shot with artifacts known to be the product of intelligent agency, how can we have confidence in what it tells us about artifacts that we don't know about?
Well, actually, with regard to #2, it wasn't until #9 that I caught on to my own stipulation that design can mimic chance, the criterion being useless for ELIMINATION of design; and only useful for DETECTION of design.
After all, if it has what appears to be a 50-50 shot with artifacts known to be the product of intelligent agency, how can we have confidence in what it tells us about artifacts that we don't know about?
I actually agree with you here; that is indeed the question. I tend to think that if there is a reliable test then there is a lot of research and refinement that needs to occur.
I tend to look at the problem like a murder mystery, sort of like the way we could tell whether Laci and Conner Peterson were murdered, or whether their deaths were accidental. If we were on Scott's jury, would we have any credulity in his tale of an amazing coincidence that he just happened to be fishing in the same general locale where his wife and son's bodies were found, but that he had no part in their deaths? The question I keep asking myself is how to quantify something that we intuitively decide all the time, which is, how we determine whether something is accidental or caused by an intelligent agent.
Cordially,