It's simple. First of all it is pretty much disclosure that they can not fight backlink spam and can't solve that part for years. Especially when it all started to be all about money "crawling budget etc..."
Mueller continues:
They have most answers already to all the questions that they have scraped from all over the web. They are not crawlers pretty much any longer, but scrapers IMO. No better than us trying to swim out there in these waters.
To understand more what I'm talking about, this patent submitted by G recently can shed some light to that:
https://www.seobythesea.com/2022/05/generating-query-answers/ I'm in no way related to that website.
This is why you get straightforward answers to questions that have static answers. They don't need any more sites to reply that questions. This is also why insane traffic drop in PAA and more sites. All related to that IMO.
So they don't need to recrawl any site containing that information. Again, saving money.
Crawling inner site pages will be fine, once when they have website in their db. For all other actions they will ask for submission via GSC or live indexing API and take it from there. Again, not recrawling junk, just acknowledging it for the first time.