I see your point but I still think you're being unfair. Viliam_Bur's hypothetical leftish ideologue reads to me as only a little more strawmanish than his hypothetical libertarianish ideologue. I can imagine a right-libertarian Multiheaded from a parallel universe reading VB's post and complaining (for example) that "I expect government to introduce bureaucracy and pervert good ideas" is a terrible strawman on the grounds that all right-thinking right-libertarians are well aware that using "bureaucracy" as a derogative is an unreflective cheap shot, that government bureaucracies sometimes merely substitute for private bureaucracies, that governments have implemented good ideas without perverting them, and so on.

Admittedly, when I read Viliam_Bur's comments on politics, I do occasionally suspect the unconscious operation of the thought process you sarcastically identify (which is not to say it's ever the dominant thought process at work). In fairness, I think VB himself often detects it too, and tries to consciously offset that in-group sensibility bias when he notices it. But now I am piling one speculation on top of another.

In any case, I agree that Viliam_Bur's meta-level argument here is correct: ideologies make conflicting empirical claims as well as conflicting normative claims, and so empirical claims can reveal as much information about one's ideology as normative claims can. For a local example, look at the correlation between politics and expected probability of global warming on the last LW survey. (For an even more local example, see this thread.)