The remaining 5% could in a normal, air-operated concentrator be `anything else' that was present in the air. Looking at this example and explanation, anything that ends up in the outlet stream is what was not absorbed. Then the question remains: does absorption occur equally for all air components (except $O_2$ of course) or does it differ?
I can give you some numbers (see also the presentation I referenced above), but the absorption depends strongly on the type of zeolite. For example this $AgA$ zeolite has a 1.63:1 $Ar$ selectivity and a 5:1 $N_2$ selectivity with respect to $O_2$. Whereas the $LiAgX$ zeolite has only 1.1:1 $Ar$ selectivity.
In this article they mention that they can get 95% $O_2$ with the remaining 5% being almost completely $Ar$. However, they use a contaminant free inlet mixture of $O_2$, $N_2$ and $Ar$.
The most interesting for you is probably this article. They study how $H_2O$ and $CO_2$ affect the operation of a zeolite oxygen concentrator. What they show is that there is not going to be any $CO_2$ in the outlet stream, but instead $CO_2$ adsorbes so strongly on the zeolite that it will degrade its overall efficiency.
In summary, to answer your question, species like $CO_2$ and $NO_2$ will deactivate your zeolite, but this is due to excessive adsorption so this means that the outlet flow will only contain $O_2$ and $Ar$ and sometimes a small amount of $N_2$.