Yes, of course, because "everyone from [insert country here] is [insert generalized comment here]" has always worked out well in real-world analogues. The definition of culture is dependent on its source - do you want the anthropological definition of culture, or the political definition of culture? Also I feel obligated to point out that a "flaw" in a set of beliefs is based in perspective - what I might say is a good value you might see as bad, or vice versa.'
Over-universalization of culture is an issue when expectations are made that every single member of a society believes X and so warrants an action as a consequence: it's shoddy excuse for collective punishment. Under-recognition of culture, however, can be just as disastrous: in not understanding the common and dominant viewpoints that exist in a region, excessive commonalities are perceived between the evaluator and the subject while real and common differences are neglected.
On the harmless end of the spectrum, the under-recognition of cultural difference blinds many people to facts like, say, Thedas has no philosophical concept of cultural practice analogous to Western Liberalism. That's just relatively well off westerners (the primary Bioware demographic) judging a fictional midievel setting through 21st century liberalism.
On the harmful end of the spectrum, people who expect belief systems be shared or appreciated group together incompatible cultures by drawing poorly thought lines on a map or try to remake distant and different societies to their liking. And of course that has never gone poorly either.
Of course flaws are a matter of perspective. All evaluations are a matter of perspective. I feel no shame in having a perspective that evaluates, say, a racial narrative that has a highly selective depiction of history to cast a one-sided grievance narrative is a poor basis for a healthy society.
As for a working definition of culture, I'm fine with 'the learned and shared behaviors and perceptions of a group which have been transmitted from generation to generation through a shared symbol system' if you are.
Collectivization is a term used to refer to collective farming, so I don't really know why this is being brought up in your argument. I'll assume you meant collectivism, in which case there are reasons people are so opposed to such a mindset. For one, it stifles individuality, which is why I would never use it when I am arguing for viewing Dalish not as a whole, but as individuals with differing opinions. There are a million other arguments I could make against it, but I doubt anyone wants to hear me ramble on and on about statism. Not that statism can apply to the Dalish, as they have no central authority. Hence why collectivism cannot even apply to the Dalish. Perhaps you could clarify what you meant? I don't really grasp what you're getting at.
Collectivization is a term I have seen used in a wide range of topis, not just collective farming, but whatevs. You don't want to use collectivism either, though that's probably because you're thinking of it in the statism context. So let's use macro-level and aggregates.
Grouping individuals as aggregates is used and useful on macro-level analysis because, for all intents and purposes, individuals don't matter in determining the common, dominant, and accepted viewpoints of a group. It's the collection of lots of individuals, and while each individual of a group doesn't necessarily reflect the views of the entire group the consensus views of the group will be reflected across most of the individuals (or else it wouldn't be a group).
The most common western use and reliance of aggregate analysis is the democratic election poll: not every person votes for a particular party, and not every person who does believes all the tenants of the candidate/party they are voting for, but the the candidate with the highest vote is generally regarded as both a legitimate and reliable reflection of the group's accepted (or at least most tolerated/least opposed) interest, with the strength of the voting reflecting the strength of consensus and support.
Cultural recognition works in similar ways, and can literally be the difference between life and death if it is ignored or not. In the west, a left handed male offering his hand uninvited to shake a woman's is uncontroversial. In Afghanistan, it can be offense enough to warrant changing position and supporting attacks on the perpetrator. This could be predicted had dominant cultural beliefs been considered and followed: the left-hand being the dominant ****-wiping hand, and the cultural considerations of what is considered respectful to women.
Like any type of modeling, it is not and never should be treated as infalible. But modeling off of dominant patterns and trends is a far more reliable way to interact with cultures than not going off of the most common beliefs and practices- and, with awareness and ability, you can further calibrate to adjust to sub-cultures, and then sub-sub-cultures, and so on. The understanding of groups works from the top down, from the most powerful trend setters and dominant belief systems to the least, not the bottom up (where the least relevant beliefs provide the most signal static).
Of course both sides are, to an extent, filled with individuals of such a mindset - I never wrote otherwise. There's really no such thing as "clean hands" in the Dragon Age universe, now is there? It doesn't mean you should generalize either side, though. Especially considering laws change through evolving standards of decency.
It's when people over-generalize without awareness to the limitations, and when they use their generalizations as justifications for harmful action against a collective, that it gets problematic. Generalizations remain useful, however, and downright necessary when addressing collective groups. Even the concept of 'the Dalish elves' is the use of a generalization.
As for the rest, I certainly agree. There are no clean hands: it would also behoove the people involved to realize it as well. The Dalish narrative of history is far too one sided to help: when the dominant Dalish understanding of the fall of the Dales is as a 'betrayal' because of 'religion', while finding any reference to Red Crossing or the Dales (in)action in the previous Blight or any justified human grievance is harder than... well, harder than finding recognition of the other side's grievance, it's not a helpful cultural bias.
I have high hopes for Solas to break the mode, though.