Opinions and actions are spread across continua. The ones at the ends are sometimes called ‘extremist’, ‘fanatical’, ‘fundamentalist’ or ‘zealous’. These are insults or invitations to treat the supporters without seriousness. Other times the far reaches of a continuum are admired as ‘sticking to one’s principles’, ‘consistent’, ‘loyal’, ‘dedicated’, ‘committed’. Claims of certainty and crossing your heart and hoping to die are also looked well upon. So what’s the difference? Obviously the correct answers to some questions are at the ends of spectrums while others, such as optimal trade-offs, tend to have more central values. Is this what determines our like or dislike for centrism and extremism? Lets look at some examples from my understanding of popular opinion.
Things you should be extremist on:
- What’s the worth of a human life?
- At what degree of temptation should you cheat on your partner? Break the law? Break a promise?
- How long should a marriage last?
- How much does average IQ differ across races?
- How much should a pedophile be willing to pay for you to let them have a child?
Things you should not be extremist on:
- How closely should we follow a single political principle, such as libertarianism or communism?
- What proportion of situations should you analyse in terms of a single theory?
- How much of your sacred text is literally true? How much should it influence your life?
- To what degree should one principle, such as utilitarianism, define your ethical views?
- To what degree should you rely on reasoned thought for opinions?
- How much of your time should you devote to a single activity (with the exception perhaps of looking after your family)?
I can’t see that the first list contains fewer trade offs than the second list. In fact it probably has more. So what’s the pattern?
The one I see is whether commitment is to an impersonal idea or to a group or person. If you take a centrist position on your personal and group loyalties you are something between flaky and treacherous. You are not supposed to trade off friends. On the other hand strong commitment to a policy position, theory, type of analysis, ethical standpoint, or other impersonal influence on behavior is unbalanced, biased, radical, dangerous, and consists of seeing everything as nails. It’s worse to belong to an edge political party than a central one, but worse to be undecided (central) on which group you belong to than to pick one and support it loyally.
This seems to make sense evolutionarily, as it is important for humans to have loyal associates, and not important for them to have associates who are committed above all else to something abstract that they might sacrifice your welfare for at any time. Ideas do not have babies with you or share their mammoth. Ideas are handy of course, but you want your associates to use them flexibly in the pursuit of upholding their social commitments, rather than using their social commitments flexibly in the pursuit of other principles.
What about sticking to one’s principles? That seems a praiseworthy non-human related extreme. Can you be praised for sticking to any principles though? No. Principles about loyalty, compassion, and honesty are good for instance, but principles like ‘always work when you can, regardless of what your wife thinks about it’, ‘always walk on the left hand side of telegraph poles’, and even committed utilitarianism impress few. Again it’s all about absolutes of reliability to others.
Meteuphoric 
So the one kind of extremism we admire is extreme loyalty to associates. All else is dangerous.
The label extremist, in everyday discourse, is not always applied to those who holds beliefs at the far end of an evenly distributed contiuum – even though that is perhaps the most common sense interpretation of the word – so much as the not-quite-identical notion of people who are extreme merely by virtue of disagreeing with with someone else’s strongly held opinion (possibly, the majority’s) on something considered important. By usage, it is close to being a (more derogatory) synonym of “radical.” So someone who thinks Star Wars is the greatest movie of all time is no doubt close to the far edge of a continuum, but would be unlikely to be called an extremist – a fanatic, perhaps, but in a mild sense.
Strength of commitment to an idea should also not to be conflated with where the idea lies on a hypothetical continuum, even though there is a relationship. I am very, very certain my bed exists, which most people would find uncontroversial; other friends of mine are quite certain God exists or doesn’t exist, but few contemporaries would describe either camp as extremists (even though these are radically different beliefs.) While I advocate significantly increased immigration for this country, which puts me a long way from the median opinion if polls are to be trusted, I hold the view reasonably undogmatically, or at least I hope I do – I can easily envisage the kind of arguments and evidence that would change my mind. Some people would no doubt call my views extreme, while others might not; loaded words like that are often used very equivocally.
I don’t agree with your assessment of popular opinion. While most people would probably hold strong, unequivocal, and similar views about the value of life and about handing children to pedohphiles, I don’t think many would advocate “extremism” on your other three points, and in fact I’m not exactly sure what such extremism is supposed to constitute; nor do I think there is much strong popular consensus about them.
I also take issue with your second list. Quite a lot of people advocate single-minded pursuit of a goal, spending a lot of time on it. Most people would agree in principle you should use reasoned thinking absolutely as much as possible, although would no doubt then backtrack or take issue when confronted with their own irrationality. Application of a single theory to multiple situations is far too vague a notion to engage with one way or the other; it depends critically on what sort of theories and situations you are talking about. I happen to believe strongly in applying the Intermediate Value Theorem all the time in Calculus, and Newton’s Laws of Motion most of the time in the Physical world; but I recognise that say my political theories are not quite so rigorous and aren’t always a good fit to the complexities of reality – social science is largely underdeveloped, and in most cases solid bodies of empirical evidence aren’t in. This is perhaps behind most people’s fail to be “extremists” on single political ideals like communism or libertarianism – also the fact that the historical record has seemingly lead to bad pragmatic results for them, although you might argue that modern political systems just represent those instances of radicalism that worked – the successful mutations if you will.
Obviously since I don’t accept your data, I’m not persuaded by your inference from it; although even accepting the premises I still don’t buy the line your argument, since a lot of cases of group loyalty arise from mutual commitment to abstract ideals. In fact I’d wager your average local Communist Party meeting or Pentecostal Church service would evidence extremely high degrees of ingroup loyalty – more than say nearly all people’s loyalty to their race, which is what I assume you are referencing with regard to IQ variations – and that for most extremists this is at least as big a motivating factor as the abstract reasoning that initially leads to the association.
I’d accept we’d likely have more psychological predisposition to be loyal to people than to ideas. But sticking to one’s principles is certainly looked up to, in many contexts at least. Look at Malcolm Turnbull – even his enemies, while despising his style and vehemently disagreeing with his views, were privately expressing grudging admiration at his refusal t0 back down.