This argument is against patenting 'junk' DNA: but European and American patent law both require a patent to have industrial application and to be novel.

So, I find it unlikely that this situation should apply in practice (although it is worrying that it might apply). And, I don't find it an argument against patenting useful non-junk DNA.

Note: I'm probably on your side, at least in that I dislike implementation of patent law in general.