How true to do suspect the following hypothesis is: FAANG overpays for top talent they do not need purely to keep it out of the hands of startups and competitors who do have a present concrete need for the top-shelf talent pool?
I don't know to what extent that's an explicit strategy, but I would find it plausible to imagine that FAANG could spend a lot less on engineering - both in terms of salaries and head count - to keep their current velocity on product development, but they see a strategic interest in controlling that talent pool which they have an opportunity to exploit via capital advantage.
The other side of this is mentioned in the OP - at scale, if you get top talent for $$$ and can squeeze 1% more revenue out of your products, it's still worth your time.
