The front page says that this is a site
for linguists, etymologists, and serious English language enthusiasts.
And I disagree with that.
No matter who you would like to be for, or who you should be for, it is not who you are for right now.
I spend a lot of time on Cooking, and know fairly well how a SE site works. I have also asked questions on lots of the other sites of the network. Each time, I faced the challenge of determining if my question is a good fit to the community, or would be frowned upon or closed outright. With time, I noticed that this was commonly determined by just a handful of users. These few users were very active, and they steered the content in certain directions, even if they weren't moderators. Stated this way, it sounds a bit questionable, but in fact, they were no dictators. They were just the high involvement users, who had taken it into their hands to take care of things, and the community was glad to follow in their lead and to support their decisions with abundant upvotes.
Obviously, these users are the experts on their sites. They aren't always the professionals who get paid for doing this in their daily life, but I nevertheless consider them experts. And the good thing about experts is that they can distinguish quality from superficial attractiveness. Of course, the democratic system allows the crowd to push polls and other unwanted question types to popularity. But I noticed that the trick to find out which questions are considered good content for a site is to find out what the top users see as high quality.
Together, the top 13 users of EL&U have delivered 4800 answers. This is 20% of all answers on the site. This seems like a good tradeoff between coverage and a manageable number of users to check on. I couldn't see what they have voted as high quality, but I had a (probably better) metric: I read through their favorites. And sure, some interesting patterns emerged.
First, these users tended to favorite questions which weren't terribly popular. While they had attracted an above average number of votes, there weren't the kind of questions with dozens of upvotes which smell of a "me-too-question". This supports the hypothesis that these users choose quality questions without falling for the usual bias. Then I noticed that the level of question complexity varied. There were definitely entry-level questions present. This was a very good finding. We want to know what type of questions is good here; the consensus on most SE sites is that beginner level questions are welcome, if they are of the right type, and one of the supermods confirmed this recently for this site in chat. Third, there was some nice consistency within the set of questions of each user.
There were four users I didn't consider, because they only had two or less favorites, not allowing me to see a pattern. The rest had 5 or more favorites. Two of the relevant users had a preference for questions with a scientific orientation. They liked discussions about the natural rhyming patterns in English, about the presence of certain sounds, or about current trends in language. In contrast, the choices of the majority of the users were dominated by hands-on everyday questions on usage in very specific cases. These were questions asked by practitioners with a specific problem at hand. Some were at a level which even non-native beginners would find easy Is it wrong to say "cook a cake"?, and some sounded sophisticated: How should I pronounce "Worcestershire" as a rhotic English speaker?. But even when they had some linguistic relevance (it is probably hard to find a question about a language with no linguistic relevance at all), they were definitely written from the point of view of a practitioner trying to solve a certain problem.
I am aware that the availability of these kinds of question could be causing a bias in itself. But first, we have the evidence that top users with interests focused elsewhere still found enough questions of the scientific kind - they were there if one wanted to look for them. Plus, the large share of practical questions is circumstantial evidence that the community finds this kind of questions interesting and useful.
My conclusion is that, whatever the community has noticed it or not, it has already defined what it wants to see on this site. The top users aren't everybody, but they are the ones everybody (mostly) agrees with, thus forming a reasonable proxy. The fact that the "decision" is not made by self-reflection but by everyday behavior makes the result even more valid. We who study people know that they often don't know what they wont or need, but their behavior and reactions to the environment is a reliable predictor for their needs. So, the linguists here are definitely in the minority, and experts and questions on "applied language" are needed.
I don't say that linguists and their topics should be crowded out of the site. They provide valuable input, and without them, the content would be poorer. But the priority should lie elsewhere, and there is no need to try to decide whether a question is a bad fit for the site just because it isn't scientific-y enough. For good or bad, this isn't a site for language researchers, just like SO isn't a site for computer science researchers. It answers the questions of those who practice English everyday without making it the object of study. If we want to attract experts, we should probably aim for this class of people. I don't know who they are - probably journalists, editors, writers and everybody else who is judged by the quality of text they produce or just takes a pride in it. Don't forget, the linguists and etymologists may be somewhat more marginal here, but for a community, you definitely need the enthusiasts.