Yeah, its one of the interesting consequences of gamifying reality. D&D's goal isn't to be a life simulator, its goal is to be close enough to one while still being a game. So you get trade-offs like this.
One of the elegant things about advantage/disadvantage is its a fairly simple system that gets rid of a lot of bookkeeping around tracking modifiers. Now a lot of that bookkeeping was there to simulate real world things like targets being harder to hit when you can't see. And advantage/disadvantage still does that, but it does it in a way that's easy to remember and easy to track. So of course when you run into cases where a lot of things use the advantage/disadvantage system, things might get weird, but it still works in a way where you don't need to calculate 50 different modifiers in order to do something cool. Now, of course, if you like calculating those modifiers, there's nothing wrong with that, but it does take quite a bit of time at the table when you could be having fun describing your awesome shot at the dragon from 300 feet away.
I actually was in a game with a similar situation recently to your initial example of firing at long range and the reaction at the table wasn't "this is unrealistic that the blizzard and the long range can be cancelled out by hiding" but "I'd like to hit the dragon but I have disadvantage. What are things I can do to solve this problem and is it worth doing those instead of shooting at disadvantage". Because the trade-off worked for my table, but I can totally understand if it doesn't work for yours. The one area I don't like stacking advantage/disadvantage is the variant flanking rules. If I used flanking in my games, I'd probably use a flat bonus.