This is the point where more technical information is needed because there is a chance (and I've seen that deployed across larger enterprise system) where that information gets anonymized throughout the process before it's stored (and you still limit access to it through principle of least privilege and so on.
That's why I think it's critical for Apple to comment on this before everybody jumps to conclusions. I'm not saying they are innocent in this but making conclusions based on certain data points on older iOS is far from ideal.
Of course but Apple is a fairly opaque company when it comes to security stuff. They say "trust us, we're looking out for your best interest and not our profit" and most people are happy to do so.
Whether you trust them to anonymize a data report originating from your device
after they receive it, when they can clearly use it to identify you, even when they say it cannot be used like that, is up to you.
iOS14.6 is just a year and a half old. Plus, these guys say it's likely this behaviour persists in iOS16.
You can ascribe malicious or incompetent behaviour as seems more appropriate to you. I personally go with incompetent unless I have reason to believe otherwise. I mean, at some point iOS code completely failed SSL certificate validation (as in, validated anything) because of an if fall-through, which to me was interesting not as much as the bug per se, critical as it was, but the complete lack of regression testing in the SSL certificate validation which should have caught such a bug in pre-release stage. This indicated at the time a very shoddy software development process, especially in respect to sensitive security code.
If you know One Punch Man, there's a character called
King who's widely believed to be the strongest man alive and an extraordinary super-hero, when in reality he's completely ordinary and a coward. But he's got a very serious look and he's almost always silent and this lets everyone project their imagination on the character. It's extremely funny.
I feel that Apple's security and privacy reputation are based on obscurity on their side and projection on the user's side.
And someone could easily access this personal contact information by getting an iCloud identifier like those they found?
Well, Apple can for sure. They say that analytics data are not personally identifiable (I quoted you their own document) but this appears to be incorrect. Apple has everything it needs to make the connection, since they send analytics reports with uniquely identifiable iCloud ids.
I understand this is not an issue for you, perhaps you'd be happy to share analytics which can be traced back to you (most people wouldn't care).
It doesn't mean it's a non issue

Your logic is a little circular: you trust Apple in respect to privacy, but when it is revealed that something Apple does is not as they say it is, you nevertheless don't think it's an issue because you trust Apple in respect to privacy.