The potencies of these benzodizaepines at these receptors is very weak compared to their potencies at GABA A receptors.
The paper even mentions in their discussion that this is not phisiologically relevant except for when high doses of benzos are injected into the spinal cord directly.
I would reccomend reading the discussion sections of papers, as while there is often a tendency to overstate results, papers will usually at least attempt to provide plausible limitations for their findings.
Reading only the abstract of a paper robs you of the context and honestly helps propagate misinformation that is pretty common on forums such as here. This problem is honestly super widespread as it is hard to infer context without experience in a specific field, and requires effort.
If you can't access the papers try
Sci-hub (you often have to Google search for mirrors as the site moves quite a bit) or the
mutual aid science community. If unable to find papers still, ask here and somebody with access (including myself) can likely help you.
At high enough doses, no drug is selective. The true mark of "selectivity" for a drug is the ratio between it's affinity for the receptor that produces it's main mechanism of action, and a different receptor. You can think of this ratio (x) functionally as "you will need x times an active dose to produce activity at a desired target". Nonselective drugs have these ratios in the ones to tens, ratios over a thousand fold can be considered as selective for one receptor over another, as a dose that would act on a minor binding partner would likely provoke a fatal overdose from the major receptor.