"Trying to edit something" is a bad idea these days - the content is far too developed for that. Instead, post on the talk page with a description of your suggested edit, wait for people to object/complain/discuss, refine the text on talk if needed, then edit. It's pretty indistinguishable from a "pull request" workflow on github.
Yes, I know that a lot of novice-focused content advises people to edit directly. That's fine for fixing typos and bad grammer, not for substantive work.
Neither is true for Wikipedia. There are no clear owners for each article so the editors are more or less hidden in the shadows (from the readers and casual contributors, at least), instead of being front-and-center of an article and explicitly taking responsibilities. It is also very unclear why they get to “own” the article like a GitHub repository; are they the initial author? Major contributors that took the baton from the original author? In many cases they are, but Wikipedia does not make that relation clear, so the editors look like self-appointed to an outside eye, which is why people have issues with gate-keeping (not the gate-keeping itself).
Making things worse, an “outside” contributor also has no chance but to work with those (seemingly) self-proclaimed maintainers; only one article can be under an entry on Wikipedia, your own modified copy has no chance competing with the original, even if you managed to make other people aware of it and agree it’s objectively better. That’s competitive imbalance, another trait of undesired gate-keeping.
I’m not saying what the Wikipedia community is doing is wrong. But if that’s how they want to do things, they need to fix Wikipedia to reflect what they want the process to work. The current system is broken.
Wikipedia recommends doing things, not talking.
Articles are also graded, I don't know what article GP is talking about but I'd say anything below "good article" is fair game.
That's absolutely not how it's been presented for the last couple decades, though.
Second if this is the way to go - and I don't agree [2] - the UI should reflect this workflow and not propose a different one that mostly fails.
[Edit] PRs also assume some kind of owner and committer relationship, which might be right for open source projects but is not the case of Wikipedia - even if many editors on Wikipedia consider the content "their own".
[1] https://en.wikipedia.org/wiki/WikiWikiWeb - "Cunningham's idea was to make WikiWikiWeb's pages quickly editable by its users, [...] . The usefulness of Wiki is in the freedom, simplicity, and power it offers."
[2] Wrote a quite successfull wiki 20y ago called SnipSnap, the engine of which started Atlassian Confluence, and did several years of research in using Wikis back then.
This is something they should work on. But as every club that get successful and was open (I attended many Wiki/Wikipedia meetups 20y ago) it gets more restricted and close to keep the masses out. What a shame.