I think you may have missed my point in that:
The only evidence you have regarding DNA is the paper report that you received from the testing lab. Therefore the source must be a copy of that report. If you use a copy of the report as a source, then you should qualify that the source and its quality, from good to bad, as a source of that fact.
I agree that paper records are not always good, but neither is a DNA test. It can return a value that an individual was less than 100% chance of them being a family member. Or the test could be faulty, contaminated or inconclusive.
So here is what you have.
A potential relationship of two individuals A and C
You set up a relationship in the software that describes the relationship and you provide all sources that indicate the nature of the relationship, (bother, sister, father mother, uncle, grandfather, slave, godparent, girlfriend, etc.
A <==============> C Is the relationship
The you provide all sources that either verify or refute the assertion of the relationship. These sources could be letters, church or government documents, DNA reports, audio/video interviews, books, etc.
A <===+=+=+=+====> C Is the relationship
......S,S,S,S Are the Sources (+ are Citations)
Ten you evaluate the source and within the source_citation tell the world what you think the value of that source is 0-3.
A <===1=3=3=0====> C Is the relationship
......S,S,S,S Are the Sources (0-3 are Citations)
If the DNA test source is trustworthy you give it a "3" if the bible account is hogwash you give it a "0"
If you don't have a DNA test "as proof" you can still have a very reliable source that verifies the relationship, or if DNA can't be used to verify the relationship, you can still have a "3".
Secondly, what if you reader think DNA testing to verify 2nd cousins is too unreliable in a small community. Should you say "Proven by DNA"? The test may not really prove anything, but the account of the doctor noting a particular birth mark may be better.