Trusted/senior users (narrow focus)
We already operate hierachies of editors - for example at a simple level a disruptive or biased editor and a skilled FA writer are treated completely differently. A registered user or admin can make content edits that less trusted editors cannot. I know that's not what you mean, but undiscriminating acceptance of all edits and all editors is a problem when you look at it. It gets quality to a good standard; then it starts to hinder not help. In hindering, it discourages the skilled editors who can/do work on high quality -- and this is its far more problematic effect. As a high quality reference work we have to be willing to say (as Yaroslav well words it) "degrading quality to an inferior level is not acceptable".
How to achieve that? The key is not to lose crowdsourcing, but to tame it, and to ensure it cannot act negatively (even by accident, agendas, or unskilled good intention).
We're designing for the mainstream of articles, where "erosion" is an issue, poor quality edits happen, and a user who is capable of taking it to high quality is not readily thrown away because "anyone can edit" must always trump "high quality content".
Ideally we want both. The philosophical solution is to have "anyone may edit, but... if you want to do some kinds of edit, you need to demonstrate you are a safe capable custodian of those edits". It's still open to all, but a user wanting to edit some sensitive, edit-warred, or high quality content may be held to the talk page rather than allow article disruption, POV warring, or inadvertent poorer quality into it, until they show they are a safe pair of hands to edit and usually won't cause such quality issues. It's the same principle as drives adminship - anyone can gain access to some potentially risky/damaging areas needing judgement and quality, but they need to show they have community trust first.
Most articles are lower quality. We can educate and inform, mentor and guide, provide tools, and improve them. But for some issues and areas, it helps to formally recognize users who are of this kind.
I don't have a problem with suggesting that. It's in line with Wikipedia's core philosophy, and a vast part of the mainstream community wants us to be credible as a high quality reference work.
Communal agreeing/consensus that a number of users are trusted not to edit war, trusted not to edit with bias, trusted not to indulge in COI or inflaming disputes, trusted to speak to source evidence not personalities... this extends the community's ability to be assured on a range of issues, and address a range of edit wars. It provides a visible desirable quality standard any user can reach (not needing technical skills, just attitudes), and once obtained most will use well, teach to others, spread wisdom and high standards, and be reluctant to lose through indulging in bad editing.
I cannot think of a better way to spread high standards than to recognize those that have it, and make a distinction others will seek to obtain too.