I would like to open a question that would lead to how we should encourage sourcing and what the hierarchy of source material with something like first source materials (taking the US as an example: government sourced documentation, case law, legislation, administrative rulings and EOCs, government sourced research) on the high end of quality, with no sourcing or maybe wikipedia or reddit threads on the low end as a method of quality control. Given that previously sourcing was upvoted quite extensively, but we never provided the community any standard around that, we could address this head on. Think it would help with some of the weaker answers we have been seeing.
"first source materials (taking the US as an example [...])".
You might have had a point in US history at some point or another... Maybe... To be frank I'm not even sure when that might have been -- given how corruption, cronyism, and racism permeate US history.
In the age of Trump though (to say nothing of Kavanaugh and Gorsuch) you've an extremely strong headwind before you to make the case that "government sourced documentation, case law, legislation, administrative rulings and EOCs, government sourced research" are better sources than generally agreed upon (by scientists/specialists) peer reviewed articles, or the newspaper and wikipedia articles that echo them.
should [we] encourage sourcing?
Encourage: yes. Mandatory: no.
what [would be] the hierarchy of source material?
That's impossible to define, as it is in the eye of the beholder.
It would quickly devolve into whether Fox News is more or less reputable than the New York Times or not. With lots of examples from both sides that the other site is not reliable.
Wikipedia tried this. We can easily agree that on political topics, Wikipedia is full of rule lawyers that will revert any edit they do not like. With the result that on hot topics, Wikipedia is totally unreliable in practice. It is rated as 'low' by the OP for a reason.
So let's not go that way.
With respect to "peer reviewed articles," there are lots of problems as well. On top of that, there is a Replication Crisis - especially in the social sciences, which includes political sciences - because many results reported cannot be validated by independent research.
All in all, I don't think we will be better off with a formal rule.