I'm not defending Apple at all (I'm a developer, I've not specifically worked on apps for iDevices, however I've worked closely with those that take our platforms and tie them into those apps at many companies and startups)
I've also not finished reading the article, however right away, based on what I've read, I see the problem. Apple reviews a thousands upon thousands of apps a day, and the folks reviewing those apps (and now, AI) aren't going to continue to review an app when they see a potential violation. It is up to the app developer to be clear and concise.
Apple is a pain to deal with, I get that. However, I see why it was rejected. Continuing to read the article now.
EDIT: read the rest of it. My take is that they are probably employing AI based upon the odd response (which is dumb, also note, Apple has been using forms of machine learning for years to assist with this stuff. When I say AI, I specifically refer to LLMs).
Something that I think folks really need to pay attention to, however, is the first impression. The app reviewer doesn't just hit a reject button and call it a day. They take notes. Notes that aren't shared with you, just among folks that may review it again in the future. When it is re-reviewed, those notes are taken into consideration and given a ton of weight. That is why it is important to give a perfect presentation the first time around. If you have 30 seconds to re-review an app that was rejected and you have notes (and possibly a screenshot or two? I don't know if Apple does that part) of what was presented during the previous review, you aren't going to do a whole bunch of investigating if it looks like nothing has changed. You have deadlines that impact your performance review, you will move on and rubber stamp the re-review. The next level up will take slightly longer, etc.