Still a NO
It's been about four years since I first answered this question in the negative. I deleted it (for reasons I no longer remember), so am going to try again. I've reviewed the query (totally not coincidental to JBH's recent review of his own perspectives!) and all the responses and some of the related questions. I'm still not in favour of opening WB up to purely real world questions. I also think the "considerations" are a little weakly worded and will not actually help.
Why we're here.
First, foremost, and always, this forum is focused on the making of fictional worlds. That's our foundational raison d'etre. It's the one concept we should keep in mind whenever considering any kind of rule, policy, or custom. If a rule or policy change does not promote worldbuilding as an art, an activity, a concept within our community of querents and respondents --- let alone in wider SE community and beyond --- then it should be considered a bad change. All questions here should thus share this focus on the querent's world as it is in progress. Real world queries as such shift the focus from the fictional to the real and thus away from the world being created.
I agree with the principles (under Reasoning) that we have to understand the real world in order create the unreal world. However, Worldbuilding is not the place to learn that. Querents should come here with questions that deal with the application of real world knowledge --- how knowledge can be wrought into something new --- rather than questions seeking real world knowledge. The only exception to this, I'd argue, is a really good question looking for resources. This is because we do have resources here in Meta that I think we can do better to showcase.
Basically, if the question is typologically "I have a fictional world; please tell me how the real world works", it ought to be off topic here. We need to be focusing on why we're here, and it's all about the other world.
If the question is "I have a fictional world and I'm trying to fit this piece of real world knowledge into the fictional setting, but I can't quite figure out how the pieces go together", then that's a question we can help with, and is on topic.
Dumping Grounds
It's an old concern that is as valid now as it ever was. We've always complained that WB.SE is a place where the science forums dump their difficult queries, where everyone dumps the questions they don't want to deal with, and where people feel free to ask us to do their homework and basic research for them.
Literally, one of this question's answers (Mark's) gleefully shoves this whole dumping ground issue in our faces by outright and joyously crooning that he considers WB to be his usual forum's dumping ground:
I am in agreement with this, and am thrilled with worldbuilding accepting this stance. It also makes the site mesh better with other sites. Regarding:
Caveat: Members are certainly welcome to suggest via comment that a post may get better answers on one of the other sites in the network. There is nothing wrong with the suggestion and its always possible someone didn't know. That said, being on-topic on another site does not make a question off-topic for Worldbuilding.
For a small site such as engineering.stackexchange.com, about one-fourth of the questions on the site come from someone attempting to design a device for a story using real-world engineering principles. These questions have been ruled off-topic for various reasons, even when they have really interesting applications. Unfortunately, before worldbuilding we had no place to send those users. This made for a terrible user interaction for new visitors.
The typical route I take when seeing a good "what-if...?" question on engineering these days has been to route directly to worldbuilding. Since there is no practical application, or testing, and the initial conditions are wild, this site lends itself well to resolve real-world designs without getting into excessively complex models. Overall it is more interesting when we can use unobtanium as a potential material of construction, something which is never allowed in engineering - but perfectly reasonable in worldbuilding.
Worldbuilding solving real-world problems helps the science sites stay a reasonable resource for minds seeking real solutions to real problems. Simultaneously, this allows users access to engineering, astrophysics, physics, and other scientific models to solve unique, fun puzzles and inquiries.
With all due respect, no, it's actually not our purpose to solve unique fun puzzles that actual engineers don't want to deal with. I'm so sorry that new members over on Engineering had a terrible time in your forum, but it's actually not our purpose to entertain those questions either!
Possible Solutions
I'd argue that, if anything, the situation Mark cites is indication of a clear need for a 'FunScience' or 'Science-lite' forum here on SE. A place where folks can ask the purely theoretical & hypothetical queries that fit neither WB nor Engineering. Such a forum would remove the burden of "real world questions" from Worldbuilding, and would remove the burden of "hypothetical questions" from Engineering & Physics.
In the mean time, the only viable solution that solves both of these issues AND maintains our uniqueness of perspective is to make "real world questions" off topic, unless there is clearly a worldbuilding connexion.
The only viable way to get the two camps to mingle successfully is to allow only the minimum amount of real world questions in this forum. This could be done by providing for very clear expectations for what's allowed. Under "Considerations when asking a question relating to the real world", there has to be more than just simple context. The question really has to be than an elaborate set-up without any connection to the fictional world itself.