No, I'm not. I cited a document titled "Wikipedia is a tertiary source" in order to establish that prominently citing secondary sources (which is what makes Wikipedia a tertiary source) is established Wikipedia policy (as described in the document).

Yes, it's stated there that tertiary sources can in principle cite primary sources directly. But in practice, if you try this, you'll be accused of violating Wikipedia policy: in particular, primary sources for anything vaguely political will not be considered reliable (even though the dependence on secondary sources from the approved list is a major source of bias) and if you can't find an acceptable secondary source then other editors will conclude that the material is not notable.

> A better way for this article would be "newsletter XY reported on a poll that said ABC", instead of pointing to the poll but linking to the newsletter.

I agree; but as far as Wikipedians seem to be concerned, if newsletter XY is on the approved RS list, things that it says happened must have actually happened (and you'll only be allowed to challenge that with another source from the approved RS list; they'll say you're doing "original research" by pointing out directly that the poll doesn't actually say ABC, because that's, like, just your analysis of the poll).

Wikipedia is not concerned with truth, in that being able to disprove content in supposedly reliable sources doesn't entitle you to correct the material.