Philipp, thanks for putting this matter up for discussion. My answer below is a response to your answer.
Your dire predictions.
The four scenarios in your answer are extreme and apocalyptic.
They portray this community as a dystopian wasteland.
They assume the worst about the people who contribute to this site.
I don’t share this perception, so I must respectfully disagree. On all fronts.
A potentially false premise.
Let's start with the fundamental premise of your argument, as articulated here:
"The mostly anti-Trump audience of the community..."
and here:
”The few but vocal Trump supporters among the whole Stack Exchange community...”
Are you sure about this?
Are these assertions based on empirical data or your own personal experience?
How do you know that there isn't a large contingent of Trump supporters here, but they choose to remain silent so they aren't called "racists", "fascists" or "nazis"? For many, it's better to keep a low profile than trigger a vicious ad hominem attack.
But beyond name-calling, Trump supporters face much more serious backlash.
How do you know there isn't a large contingent of Trump supporters here, but they choose to keep quiet so they aren't:
Unless you have reliable data to back up your premise, you could be making the same mistake the pollsters did in 2016.
The 2016 polls.
Throughout the 2016 general election the polls had Hillary Clinton leading Donald Trump by a wide margin. On Election Day, here are some of the numbers from prominent sources:
The New York Times, Tuesday, November 8, 2016

The Huffington Post, Tuesday, November 8, 2016

The Princeton Election Consortium made the following prediction on Election Day:
The [Clinton] win probability is 93% using the revised assumption of polling error, +/- 1.1% ... Most probable single outcome: Clinton 323 EV, Trump 215 EV.
The polls were wrong. And they weren't just slightly off. They were ridiculously inaccurate.
A national association of major pollsters (AAPOR) produced a report in 2017 evaluating the reasons for the polling inaccuracies. They determined that a primary cause of failure was "Trump voters who participated in pre-election polls did not reveal themselves as Trump voters".
One particular pollster attributed the failure entirely to: "Non-response among a major core of Trump voters."
It's worth keeping this experience in mind when observing that a group of people (such as the Stack Exchange community) appears to be mostly anti-Trump.
This community can handle controversial questions.
Circling back to this site, there is plenty of evidence to suggest that combustible subject matter can be handled respectfully and responsibly by this community. Here are three examples:
So, again, I believe your scenarios are exaggerated and you don't give the members of this community enough credit.
Is President Donald Trump a racist?
You ended your answer with:
What I do care about, though, is that the current political climate in the United States polarizes the community into two blocks which are incapable of talking to each other in a civilized manner when it comes to certain topics.
I guess you didn't notice that my question was asked in a civilized manner and was an attempt to engage all sides in a conversation.
And if we want this website to fulfill its purpose of being a question & answer website which explains politics and political processes in an objective and neutral manner, we can not allow it to turn into another flamewar battleground.
Even the risk of pandemonium should not be enough to justify the exclusion of controversial political questions. That approach makes this site mediocre, as important issues aren't being discussed out of fear.
Maybe my question does need to be reformulated. Maybe it's too long. Maybe the association with fascism is unnecessary. I've received plenty of constructive feedback. I may post a new question in the future. (I won't edit the original in order to keep this meta discussion intact). Thank you.