If they want a change made that's basically a completely trivial detail that was already within my ability to have done when I was doing the work (e.g. just putting some padding that just felt about right), then sure, I guess. Not sure it's actually important, but whatever.
If it will go against other work (e.g. the padding comes from some central styling rules), entails substantial off-route work, or adds future maintenance debt or project risk, it should go via the product manager if there is one. A good CEO knows not to preempt the people who decide how the product works, what it needs to do and how the team(s) are going to get there. He should also know that product will consider a valid suggestion (and he should know if they're so slammed that they can't even think about it, in which case he can help in other ways than moaning about padding).
In this case it sounds like Zuckerberg is bikeshedding over something trivial (doesn't get much more trivial than padding on a specific button) just to swing his dick about. If he has something insightful to say about the business logic, that would actually show engagement and understanding.
The parallel to Steve Jobs is a fair one, except that Steve was brilliant at design and UX and got things right. He wasn't bothered about tiny details like the padding on a single button but rather the design of the entire system, and he picked up on things that didn't sit in the overall architecture brilliantly.
He also drove people to leave Apple, and that was a bad thing. Things aren't right or justifiable just because they're like things Steve did. Sometimes he got things wrong, particularly when it came to managing people.