2000 rated players of course won't get mated in 3 moves, but position after 1. e4 e5 2. Qh5 Nc6 3. Bc4 g6 4. Qf3 is not unplayable for White - Black is equal or maybe just a little bit better.

In lichess Blitz database for 2200+ rated players there are 13883 games with this position. White won 50%, drew 7% and lost 43%.

Of course this position is nothing to write home about, but for online Blitz it's good enough.

edit: what the other responder says

I'm not saying I'm mating people in 3 moves. There are many unique mating opportunities in queen openings and many ways to create immense pressure on the king's side if the player does not defend well. Try it yourself and you'll quickly see that -2000 rated players (online) struggle with it. Similarly, you're not sacrificing development and you castle at the same moment (sometimes after black).

For example, many games I would get my queen attacked on F3 and then I would lose my rook. But this 3 move knight play on the opponents side gives you a lot of moves to create pressure, in one case you can have 3 pieces developed, while opponent has 0 (excluding the knight stuck in the corner). With the queen on king's side, you have a lot of options to equalize. If opponent is focused on saving the knight, the moves are just bad.

In 2+1 and 3 min blitz it's unlikely you'll get someone that won't make the position almost even after 10 moves.

The biggest hurdle to progress was my endgame and in blitz I found that being a piece down is rarely an indicator of loss at that level. + I did win a lot on time.

I was playing 200-250 blitz games a day and would analyze many, but I never bothered with openings. I quit because it was obviously too much.

No one is saying you won in 3 moves. You didn't accidentally get to 2000 by playing trash openings, period.

Of course, it was by consistent 6-12 hour daily chess activities where most of the play was 2+1/3 blitz and 80%+ of the openings used queen very early. I played for fun without any particular goal outside of short term winning. I relied on my own learning progression and accepted it was inefficient. I'm pretty sure after 1-2 years, others could have achieved much more with the time I put in.

I am not claiming it was the fastest way to progress, that it's a way to progress forever, that it doesn't have any flaws.

Chess has a very big space of patterns that you can learn that have nothing to do with openings and in my experience they count much more early on.

I shared my experience out of curiosity. Are there others who experienced something similar?

From my experience, to progress when learning anything it's enough to be consistent. The brain does its thing during sleep. When a roadblock is reached, you have to be able to identify it and make a plan to improve. Would it be more efficient to have a tested plan or a tutor? In most cases, yes.

Sorry but I'm not believing this without any proof.

This whole post is about how difficult it is to study chess and to improve, and you just stumble to 2000 playing for fun, with no study and no endgame knowledge. Also, consistency is most definitely not enough to improve: if you consistently practice the wrong things, you won't improve. And, if you didn't study chess at all and just played, like you mentioned, I also find it hard to believe that when you hit a roadblock you were able to identify what you needed to practice and implement a plan.