Perhaps there’s a misunderstanding that I was trying to make a point, or an argument, regarding semantics or legal terminology of types of evidence in a formal discussion.
I believe a DNA match which is precise to only one single person carries much more weight as evidence in a criminal investigation than “blood type O”, which matches to over a third of the population.
Hair analysis that was first used in the 60’s in a murder conviction has now been discredited.
Going back to my original post and the point I was trying to make that seems to have been lost here; we have come a long way since the 60’s and have more, and better, investigative tools to gather evidence to solve cases and bring them to a trial than they had in the 60’s. JMO
I don't want to miss your point that LE has better investigative tools than they did 60 years ago, and can make decisions about the likely involvement of a POI with greater accuracy so I'll just center that and say yes - absolutely true.
However, two points I'd make for understanding crime (this is long-winded, but in case you want to know):
As @Ozoner said, DNA evidence is circumstantial evidence. Even really, really strong DNA evidence is still circumstantial evidence. Direct evidence is testimony that the witness saw the material facts of the crime with his or her own eyes. But witnesses can be unreliable/lie. So the direct vs. circumstantial qualification of evidence doesn't have anything to do with the strength of the evidence.
Second point: in forensic DNA testing, there is never a match that is 100% precise to one single individual. The idea that a suspect exactly matched a sample from a crime, and no one else on earth could have matched, is a fallacy. Even "slamdunk" DNA evidence that convicts people and puts them away for life, is NOT an exact match to their DNA. That's because at no point in forensic testing is the ENTIRE genetic profile of the suspect compared.
Instead, about 13-20 locations in the genome are compared to the sample from the crime. These locations were chosen because they are highly variable among human populations, so if you can sample all 20 of them, with each subsequent location where the genetic information is found to be a match, the probability that the suspect is, in fact, the contributor increases. Nowadays 20 locations are used, making the probabilities (if the DNA sample is complete) even more accurate...though still not exact.
However, even if there are matches at all 20 locations LE still can't say that there is NO ONE else in the world who might also have the same information at these 20 locations. After all, there are several billion of us on earth, and only 20 locations. Did you ever wonder why prosecutors use the term "there is a 1 out of 5 million chance" (for example)? They are reporting the chance that another, innocent person would ALSO share the EXACT same genetic information at all locations that forensic testing looks at. They are saying "this genetic information is a match to our suspect, but you could expect, if you looked at the DNA of several million people, that 1 out of every 5 million people would also match, or have this SAME genetic information encoded at these specific sites."
Now, is a 1 out of 5 million chance that another, innocent person contributed the DNA at a particular crime "good" evidence? Of course - taking into account other evidentiary aspects of the crime, this can clearly convict a person. However, is it an "exact match to a one single individual?" No. LE can never say accurately that DNA evidence, by itself, shows that no one else could have contributed the sample. LE might be able to say there is a very small probability that someone other than the suspect did.
Now, let's just say, for example, that in the Delphi case LE only have a partial DNA sample to work with. Remember, they don't test and compare a full genetic profile, gene by gene - they are looking for their 13-20 specific locations. If they only have 9 of the 20 locations, let's say - what does this do to their probabilities? Likely, it will go way, way down - now LE may only be able to say that the DNA profile shared by the suspect and the sample from the crime is expected to appear in 1 out of 400 random people. The exact probability will depend on how common the genes are in the population at large. Suddenly the DNA evidence is much weaker. In any sports stadium in a large city, you might find several people who "match" the profile from the crime. So then in prosecuting this type of partial DNA case, authorities have to rely much more on the other evidence that points to their suspect because "1 out of 400 chance" is not nearly as impactful. Is partial DNA what LE has in the Delphi case? Only LE knows.
All MOO but for an accessible overview of why a DNA match can't determine guilt all by itself, see this article: If Police Find a DNA “Match,” That Doesn’t Mean They Have the Right Suspect