This blog post specifically says that it's a walk-back of the policy announced in your link.
Have they had a fundamental turnover in management, indicating a new pro-privacy culture? Did they move their development out from under the thumb of the CCP?
No and no?
So what’s changed?
If they weren’t trustworthy before, they certainly aren’t now.
They were mostly focused on workplace meetings. If that's your focus, then most of your users are employees of some company whose contact information you have. Users with unverified identities are a corner case that you may not feel is worth trying to get right.
Thanks to the pandemic, they have millions of new users who use Zoom for personal purposes (meeting with friends and family, etc.). What once was a corner case isn't anymore. It might even be their most common type of user.
I'm not arguing that Zoom can necessarily be trusted, but I can see a plausible reason how they could have gotten to where they are on this issue.
Their userbase changed in the two weeks since they announced their policy?
But yeah, while feeling that you can't trust them because they don't know what they're doing is not the same as feeling that you can't trust them because they're in cahoots with someone (governments, etc.) against your interests, in the end they're both forms of feeling that you can't trust them.
Stop looking at the empty words they say, and start looking at their very intentional and malignant actions over the last few months/years.
Why do you believe their real intention was revealed a week ago, and not today?
Trust, but verify, yes? Well, they're verifiably full of crapola.
So why would you trust?
It’s a bit of weaseling here. They say they’re offering E2E and that might be true, and it could also be true they are sticking with their original vision of cooperation with law enforcement because they “are aware criminals use the service”.
There is no required mutual exclusion.