If the dev tools, docs and tech debt are bad and slow the entire team down, then why not take it as an opportunity to improve them?
At FAANG doing things that have a multiplicative effect across the team/several teams/entire org/entire company are rewarded extremely well and a rapid way to advance your career.
Not at Amazon because you can't write some silly "customer-centric" pseudo-press-release to justify your effort, so nobody wants to be stuck paying the salaries of teams working on tools. That is why their tooling is always bankrupt.
They would all rather spend money hiring another useless Principal Engineer instead, who will walk around doing nothing but networking with the cabal of useless Principal Engineers.
For a real FANG, yes, that is the case. Working on tooling is insta-impact.
At FAANG doing things that have a multiplicative effect across the team/several teams/entire org/entire company are rewarded extremely well and a rapid way to advance your career.