Much of what you've done at other companies, other on the Microsoft stack or open-source stack, doesn't transfer. Little of your tooling experience matters.
While you can jump from one Linux-and-Python-and-git-and-JIRA-based startup to another and hit the ground running, you'll be spending months ramping up at Google unless you've worked there before.
I'm not saying this is justified, especially when it comes to other ladders like graphic designers (where you see the same thing happening), but there does seem to be a certain logic to it.
I tend to think of it this way: Google has tens of thousands of engineers. That's an ecosystem the size of a few thousand startups, except stuffed into one company. It has its way of doing things, and to move up in this ecosystem you don't just master your own startup, you master the ecosystem around you.
That and a noticeable amount of elitism.
Say what you will about Google, but I don't think it's a very controversial opinion to say they do hire a lot of the best software engineers, and they are very good at it.
I'm sorry but that is completely the opposite.
In MLB and a lot of other sports now, every single play is tracked. Every player has rate and counting stats to measure efficiency and overall production in every category. Anyone in the world can subscribe to the tape library and watch your entire career on loop and analyze every single moment of your professional career.
No one can do anything like that in closed-source software. Infact, that is the big unsolved problem in tech hiring. No one can tell who the good devs are because your "track record", your resume is just not a reliable indicator of your ability.
One of the reasons I chose to join Google over another security-related company (that would have done little coding) was that I wanted to learn how to write software better. I've learned a lot about testing, understanding the reliability of distributed systems, writing code for maintainability, etc.
