They don't. In fact they even did an experiment and admitted some people at random, irrespective of how well they did in the interviews. Those people were found to perform about as well as the legitimate interview "lottery winners" hired during the same time period.
A lot of people (anecdotally, the majority) at Google have the "impostor syndrome", and the news of the experiment did nothing whatsoever to quell the symptoms. Now they don't know if they are, in fact, not impostors, but they do know that on average they perform about as well. :-)
Could the performance of the lottery winners have been "environmental"? That is, they benefited from being surrounded by competent people (which was, in turn, guaranteed by those people having gone through the interview process) and "leveled up" due to that?
In other words, maybe as long as you let in a small number (but only a small number) of non-performers, you're fine (which is bound to happen anyway - I'm sure there is some noise in the interviews).
Yes. Getting hired by Google is only part of the deal. Actually _succeeding_ when you're already there is much more difficult. It's a high pressure environment with a lot of very smart overachievers. Because of this it's sort of a self-fulfilling prophecy, and people who don't measure up also don't feel welcome, as it were. Since performance reviews are largely derived from peer feedback, hiring mistakes tend to be self-correcting. Most of the time, though, I've seen great people leave just because they didn't like the pressure. The amount of pressure depends on the team. The higher the profile -- the more pressure (but also more rewards, greater career potential, etc). But the general bar for what's considered "good work" is pretty high, and more uniform than in any other large company I have ever worked at.
Then there's the issue that by the time you even get an on-site, you're already very much not a random candidate. Recruiters actually do look at your track record, etc. You can bullshit there, but I don't recommend it, since references will be spot checked, and they better line up.
Google interviews are largely a roll of the dice above certain level of basic engineering competence. I.e. if you don't know the basics, you will almost certainly not pass them. But if you're a more senior candidate, Google doesn't really know how to interview you, and their interview process turns into a random number generator biased heavily towards "no hire".
They are no longer amongst top choices for top people. Alphabet might be, Google isn't. That's why they are dumbing down their interviews in the past 8 years and repelling even more top people that want to change the world and not be just another cog in the machine.
They certainly still _are_ among the top choices for top people, but they're no longer the _best_ choice for most. I can't in good conscience advise anyone to join any 70K person company. "Cog in a machine" describes it pretty well. Ignore "self driving cars" and "internet balloons" and other BS: there's near zero chance you'll get to work on any of that, particularly if you don't already have a stellar track record at some company Google/Alphabet respects (of which there are very few).
A lot of people (anecdotally, the majority) at Google have the "impostor syndrome", and the news of the experiment did nothing whatsoever to quell the symptoms. Now they don't know if they are, in fact, not impostors, but they do know that on average they perform about as well. :-)