Thanks for the comments KJ. Educational as always.
However, I disagree with you on one aspect: that DNA evidence should just be "rolled into" the normal equivalence of a judgement of a relationship coming from paper records.
The reason is a basic one of human behavior. All paper records are based on facts that should not really be assumed , but are often ignored: 1) If a woman is married, she is assumed to be faithful and 2) otherwise, that she really knew who the father was, and/or accurately revealed who the real father of a child was.
Another aspect of weakness of paper records comes from often not knowing which wife of a man was the mother of a child when a man had multiple wives. That issue can also exist when a woman has multiple husbands and a child from one father takes the name of another husband.
It seems to me the "paper judgment" of: "possible" to "almost certainly" should come from paper records and they are then proved (or depending on the test, corroborated) or disproved by DNA records.