This some scraps thinking from someone who isn't uber good at English.
The not very look like introduction
This is a piece of POV and represent solely and completely my personal opinion based on my (very) limited experience of .En & .Fr Wikipedia. The purpose of this piece of POV and OR is to give an additional perspective with no pretense to be the "best way" and it won't give you solution but you may think some out of it.
It's incomplete and very badly written and ordered.
We agreed to participate to the construction of an "encyclopedia" however we never ever agreed on the definition of that very "encyclopedia". Because of that people with diametrically opposed views of what is "encyclopedic" can clash and yet both want the "very best" for Wikipedia just they don't share the same definition of that "very best".
This resulted a blurry vision of Wikipedia's aim, an erring boat that definitively doesn't know where it's going with only accidents indicating where it should not be going.
Editors have to live with that imprecise & disputed roadmap and must rely on the basics: Verifiability, Neutral Point of View & No Original Research so not only the articles but also the contents they contributed can withstand any drastic change of Wikipedia's heading.
Reference-driven content is a "very defensive" editing style aiming article & content survival that can withstand near any form of wiki warfare but the downside is that it's very constraining content wise and use article talk page intensively.
To summarize it: Only what is proved with references make into the article.
Verifiability is deadly
Verifiability is not negotiable and thus article & its contents must be verifiable. There is no Carte blanche to write whatever you want in any article.
Unfortunately issue arises due to the unequal level of scrutiny and missuses of the verifiability policy, some articles being literally shelled with Citation needed tags or worse contents simply removed by users with "aggressive" mindset.
Reference-driven contents aim to survive the worst bad faith scrutiny.
Hierarchy on the content
This hierarchy focus on the verifiability of the content and not on its relevance.
There are 3 levels of contents verifiability:
- Backed by Reliable Sources either primaries, secondaries or tertiaries. Those contents can make it into the article if they are relevant to the article meaning no trivial yet verified stuff.
- Probably true & relevant but lacking reference. Those content end in the article talk waiting for someone to prove them or not.
- The rest. That means the Original Research & other speculations. Those contents are better suited to blogs & forums discussions but not Wikipedia
This is very constraining content wise because we are limiting the contents to what we can add, rather than what we want to add.
Article creation turns into gathering enough evidences of notability as as many "ammunitions" to be used. Your article creation will be likely contested so you have to use the garnered "ammunitions" as fire support and start an artillery barrage to secure a drop zone or beach head in Wikipedia. This way of thinking is utterly wrong: clear warlike thinking, evidences of notability are used as a deterrent for would be deleter. Unfortunately current article creation is more article survival in hostile environment than happy expanding Wikipedia.
Article building & improving turns into improving the
article fortress defenses using contents as bricks and references as mortar. Having an article peppered with blue citation tags has often a strong psychological effect.
What of NPOV and not OR?
For NPOV, the constraint to only put content asserted by reliable sources force editors to harness more & better resources from the broadest selection of reliable sources if they want to expand articles. While that won't certify a perfectly NPOV article, it will at least make of it a rather well balanced article and probably achieving along the way completeness.
For not OR, if editors want to add their homebrew ideas and theories they will need to back with references from reliable sources. Managing to do so would means that it won't be OR anymore because a vetted reliable sources backed it.
New users in that
This approach is not for new users as it clear require some good grasps on verifiability and reliable sources, not to mention how to use references & citations within an article.
Reference-driven content editing works but It's neither fun, nor intuitive, nor fast. Spending hours looking for references & citations to support every single fact of an article means more time outside Wikipedia than inside and a lot of frustrations when you need +15 mins to find the right reference just for one sentence.
Editors should not use this "doctrine" of editing unless out of necessity.