Framing someone

I don't really see this particular possibility as eroding the value of DNA evidence.

It would be a pretty unusual situation in which it was possible and worth the trouble to actually fake a sample of DNA when, as noted in the article, it's likely to be easier to find a real sample of some innocent's DNA.

Even if someone did want to create a fake sample from a particular individual to whom they had no access, they'd need to get hold of a DNA database record for that particular individual, which is unlikely to exist (unless a target has been chosen from a select few known to registered with some particular database).