Generally speaking, good questions will have very few, direct answers. It doesn't really matter if there's code or, depending on circumstances, prior research, if the question can still provoke a concise answer, and even more so if there are only a few resources that answer said question; sometimes there are some obscure things that are simply hard to research, so if we get a question like that and someone knows, we should give the question a chance to get some answers. In those cases, you might suggest an edit or that the poster make an edit, but it's not necessarily worth a close vote.
Let's look at the questions you linked:
Using a flow, can you post in Chatter on behalf of another user
They admit that they know of an alternative method, but they've apparently done their research (e.g. "I know that..." suggests that they're not just shooting in the dark). Also, there's an answer that's correct, and accepted. The question may have been borderline, but it had a great answer that could help other people.
How to Auto Pause/Save Flow Interview when browser window is closed accidentally
This is a common question, and while we're not in the business of just handing out answers, if you've ever tried to research this topic, there's really nothing out there that helps. As far as I know, this answer is probably the only the reasonable suggestion out there on the Internet. Flows should be more friendly by saving data temporarily automatically, but they don't. So we need a solution.
Send public forms to your Contacts
The user is asking if the approach will work. They're simply looking for validation, which is pretty much a yes, or a "no, and here's why". There's no room for a ton of different answers. Gorav answered this one, and even included a link to a blog they they wrote about a similar situation. Excellent answer, and I don't think we can really fault the question to the point of closure (and it's now at 4 re-open votes, too, so I'm not alone on this). And the answer was also accepted. This question could use some help, but was definitely not worthy of the close votes it got. We've had questions far worse than this not get closed.
The point is, we shouldn't necessarily get stuck on the letter of the rules, merely the spirit of the rules. Our rules are meant to provide some minimum quality and usefulness to our existing users, as well as new visitors. We should definitely spend more time making edits to questions to clean up grammar, formatting, and so on, but have questions of varying quality that are acceptable in terms of provoking a few direct, high-quality, fact-based answers.
We want everyone to join us over here on SFSE, and we're not going to do that if we just close-vote anything that doesn't meet the exact criteria. We should strive to instead edit questions that are borderline, provide comments for new users to give them a chance to salvage the question, and definitely help users that have asked the same related question multiple times; they're trying to follow the rules, but they don't know what they're doing wrong, and we just close-vote them into oblivion, we'll end up being the poster children of scathing blog posts like this one.
I don't want SFSE to be the site that:
[...] hates new users
[...] also hates most other users
The reign of privileged trolls
Do we want to be labelled as a group of trolls? I certainly don't. I put a lot of thought and effort in to each question and answer, and I want everyone to feel included. You'll notice I don't have a ♦ next to my name. I'm not a moderator because I'm more useful answering questions. We all have a role here. My reputation is not a privilege, it is a responsibility, and one that I take very seriously.
Okay, so we should be closing fewer questions, and focusing on improving the quality of questions and answers. Our focus should be bringing those questions that are "just under" the bar of quality that we accept up to an acceptable level, especially for new users. That still leaves the original question: what should we close?
We should close questions that are opinion-based. Period. We are not interested in promoting one product over another, or promoting one perfectly legitimate method of coding over another perfectly legitimate method of coding. We're not here to sway the market, but to help users, administrators, and developers have a better Salesforce experience. We should always recommend doing research and not just asking a bunch of random people which is better; we're biased based on our prior experience with certain products, vendors, etc.
We should close questions that are clearly just spam as well. If they're asking about dating, selling some dubious product online, etc, we don't need that here. We're not someone's free advertising. We want our content to be relevant to random Google users and Hot Question list visitors. We're getting on that list a lot more frequently these days, and we should definitely put our best foot forward.
Before voting to close, we should simply look at the merit of the underlying concept of the question. If the question could be a good fit here, let's try and get it up to our standards, if we can do so with a minimum of fuss. Often, all it takes is a comment asking for a screenshot, code, or some minor clarification. Otherwise, if it's beyond salvaging, just hit that close link, and let's clean up the clutter.