There are multiple answers to this question depending on who you ask (meaning, what their field of expertise is).
The biological dark matter from the wikipedia article in question seems to mean sequence from metagenomic samples which was not assigned to any of our commonly known domains of life. That might not necessarily mean that it does not belong to them. Sometimes it is hard to decide where a small particle of sequence informations belongs to, because it does not match anything we know. Still, it might be from a perfectly "normal" bacterium. Or it might be a result of sequencing errors.
Additionally, depending on how you assign a taxonomy to the sequence, there might be cases where there are so many different possibilities for where this sequence came from (because it is very conserved in a lot of different organisms, even from different domains of life), that this sequence will not get any taxonomy assigned. Not because there is none, but because we cannot safely decide which it is.
Some people also speak of this information as "dark matter".
There is for example this project which specifically is designed to study the dark matter we can (or at least for some reason do) assign to be microbes, but not more specifically.
Basically, any decision of where to place potential dark matter sequence is guided by our previous knowledge about the domains of life and the genomes of all those organisms. This knowledge is very incomplete.
I am sorry this answer lacks some more nice sources. Most of the information is based on my knowledge as a person working with datasets that might include "dark matter", if I would call it that. (I call it "not assigned sequence".) I don't have the time right now to add more sources, but if needed I can maybe add some later.
As a fast add-on after re-reading your question: How much of the sequence is unassigned for example if one would study your gut depends a lot on what is actually there in you gut that we know, how diverse the community in your gut is, what database is used for comparison and how old this database is, how the data is preprocessed and analysed, and so much more. But yes, there is still a lot of genomic information in nature that has not made it into any database.