To be clear, I'm not saying that's how it should be, I'm merely describing what exists.
Technically every person has a so-called idiolect, which has a shared subset with what you would call the standard language but both includes vocabulary (and likely grammar) that is non-standard and also excludes some vocabulary (and again likely grammar) that is standard.
Just like a statistical average it's incredibly unlikely (and sometimes even possible) any one person matches that standard. The notion is only really useful when talking about groups -- as is the entire concept of a "language" (or even dialects).
The entire notion of whether something is "correct" or "incorrect" ultimately depends on mutual intelligibility and whether some authority (even supposedly "descriptivist" authorities like dictionaries effectively act as authorities) acknowledges it or not.
"Language" is a lot like "species". It's a useful concept at scale but it brakes down when you try to talk about relations between individuals and if you try to define the borders too clearly it becomes apparent that it's all just a leaky abstraction to make sense of the far too complicated mess you are trying to reason about.
Syntax rules like "no double negatives" or vocabulary like "fleek" and "pneumococcal" and "dreich"? All of that is English, and I can point to dialects of English that have each and dialects of English that don't have each.
> There are still majorities and minorities, and the first define what's "standard".
Where? In the neighborhood I grew up in, the majority uses double negatives. In the UK "colour" is the majority way to spell "color". If you want to argue that the majority determines the standards, then the standards are contextual, because different segments of the population are the majority in different contexts.