There's a certain ivory tower je ne sais quoi to it. Got me thinking about that superpermutations solution posted on 4chan leading to a paper with "Anonymous 4chan User" as the lead author.
Obviously some of these conventions might be pedantic, and criticizing an argument's form doesn't invalidate its essence, but I don't think it's a bad idea to insist on standards, even somewhat arbitrary ones. It seems like you're wishing for a world in which every citation is equivalent to "Anonymous 4chan User". I don't think that would improve things.
And even if publishing conventions are part of the problem, doing away with all convention is likely to make things worse, not better. It would amount to throwing the baby away with the bathwater.
Trading citations and authorship is such a common practice nowadays that I don’t trust that metric for pretty much anything.
Assuming that's actually true (and it's not, in my experience), that alone is a valuable thing. Do you prefer doing research in a place where political connections cannot be traced? I don't. It's bad enough as it is.
Here's the other -- really obvious -- thing it achieves: it helps you find more information about a given subject. It helps you understand precisely what a given publication is referring to.
Just because a system is abused doesn't make it a bad system. And if you want to dismantle a system, you must at least replace it with something that isn't worse.