I'm sure you're aware but, as an app developer, the review process seems random at many times.
Depending upon the phase of the moon certain features or metadata may be accepted or rejected in apps. Even when the exact same features have been accepted in prior app versions and app reviews (without changes to the actual guidelines).
In such cases, perhaps Apple's interpretation of the guidelines have changed. If so, there's no way for app developers to know this until they get hit with a rejection. Rejections rarely include explanation of the interpretation so it's up to the developer to try and interpret what the real issue is.
This is on top of the fact that, in many cases, Apple deliberately uses the guidelines and shifting interpretations to delay or destroy competitors. This occurs even if competitors are acting within the guidelines and all prior known interpretations of those guidelines.
App Store review can be random, AND this can be a good policy. (Well, good for consumers. I could see it causing developers some issues if competitors weaponize it.)
When I download a package from f-droid repository, I know the folks at f-droid have the source code of the bits I downloaded. Same thing with Debian and Fedora. Why can't Apple do the same thing? Require developers to submit machine-readable source code along with build instructions. Build them directly on Apple servers. Sign them with Apple's certificates.
I am not an Apple developer. As a consumer, I would welcome such a change.
Now, it is true that you could have a web view that fetches application chrome/content from a web server but I would argue Apple should disallow such apps anyway.
This is essentially an admission that their review process has problems.