I have come across a question which is specific to Sydney. I have also answered another city-specific question. Are those kinds of questions on-topic for coffee.SE?
I think city-specific questions may be fine but asking about stores and where to find things seems difficult because those things can change (and quite rapidly so lately because of the pandemic).
Old question, but timeless on SE. Conceptually, SE sites are intended to build a knowledge base applicable anywhere, with long-term value. There are a few exceptions, like I imagine Travel (although I'm not familiar with their scope), or say Law, where country-specific law can be asked about, and legal concepts in different countries can be contrasted and compared. But questions asking where to buy things in a particular country or city, or more generally, where to find specific resources in a certain country or city, or what's [a good / the best] XYZ in a certain country or city, aren't a good fit for the SE model.
Also, think about the follow-on questions. One of the examples in the question asks "Where to get coffee in Zurich". I don't live in or visit Zurich, but I'd be interested in people's recommendations for my own city, and maybe the cities I visit. If the example is on-topic, so would be the thousands of identical questions for every city in the world.
In addition to being too local to be of value to most site visitors, this kind of information typically has a short shelf life. Sources go out of business or become "not the best" recommendation. Or new, better sources appear, turning existing answers into outdated, bad, or incomplete information.
These kinds of questions tend to attract answers that are opinion-based, often with little detail to explain what makes them a particularly good recommendation that is more useful than the first example that popped up in a Google search.
They also attract lists of recommendations. No answer has a comprehensive list, the lists can overlap, and no answer is a clear "right answer". So the scoring system that is designed to float the best answers to the top, and to support accepting the best answer, isn't very useful.
One of the examples in the question has already been deleted. To me, the remaining example is too local and time-limited to be on-topic.
