The argument doesn’t assume this; it concludes the opposite. It claims that conclusions drawn from DNA samples can often be mistaken.

This describes the author's confusion between two different kinds of errors. We know from the premise that DNA tests can fail to distinguish between two individuals. But this would lead to false convictions, not false exonerations. The author mistakenly thinks that the error discussed in DNA tests is failure to identify two samples from the same person as belonging to that same person. This kind of error could lead to false exoneration.

The argument doesn’t generalize about other methods because only DNA tests are discussed.

The argument does not appeal to any experimental data, and even if it did, there would be no reason to believe that the data wouldn’t also hold in nonexperimental conditions.

The argument does not need to demonstrate that other kinds of evidence shouldn’t be admitted because it only claims that it’s a mistake to use a DNA mismatch to exonerate a suspect.