I generally don't think genetic research of ethnicity is very useful, it smells of calipers for mile and entire history already showed us it's basically completely irrelevant. Culture and language research is much more useful.

Majority (huge majority i think) of Russian Jews were the same Ashkenazi that were running from pogroms in Germany, they ended up all over Poland, Lithuania, Russia, etc, but there were multitude of other, non- Ashkenazi Jewish groups in Russia too - abovementioned Krymchaks and Karaites plus Sephardi, Romaniotes (Jews from Greece), Juhuro (Jews from Caucasus), Georgian Jews, Bukharan Jews, Armenian Jews etc, so even just looking at sheer diversity of those people cultures and languages the Khazar hypothesis immediately fails.

That's somewhat right, but it was gradual and did not happened in the next few centuries. Jewish diaspora existed even before 70AD, though indeed refugees fleeing from the Roman massacres speed up things. Most of Jews in diaspora lived in the Mediterranean shores, notably Egypt, but also in Italia, Iberia, Narbonensis, Greece, Dalmatia etc. where their presence is confirmed in I-II century already. Ashkenazi are decendant of those diaspora Jews that moved north together with romanisation (for example their first main community in Germany was in Cologne) and organising of states on that areas. Next was a period of opression in the kingdoms of Visigoths and Merovingian France where Jews were forced to convert or exiled, but after Charlemagne build his empire he gave Jews the merchant and financial privileges which stabilised their situation and allowed to develop into the medieval Ashkenazi in the next few centuries.

I see what you mean. Most of my exposure to the hypothesis (other than the aforementioned Zionist tropes) is from Cold War era non-Zionist Jewish sources, and they really didn't deal too much with the Yiddish thing. I believe the idea of the constant movement of peoples, in those tellings, explained why they ended up north and west of Khazar land for the same reason the Magyars and others ended ups in similar places. The main up-shot of those sources, at least in my reading, kind of goes to your final point, but in a different way. The idea being that the peoples in the Steppe were always a fluid amalgam of people and there were home-grown Jewish influences there that became a cultural seed that developed in groups in the area that sought to neither ally with the Christian world to the west and what was developing into the Persian/other empires and Muslim world to the east. So that reading of it goes that essentially no one has mythic ancestors in any one place because the version of history during any time period where one would posit a homogenous genetic group stayed "pure" from others is, at least with respect to Eurasian and African history, false. As those writers point out from the genetic (albeit, genetics as they existed a few decades ago) perspective, Jews generally are more genetically similar to the populations they live with than Jews from disparate places have genetic commonality with each other. I definitely agree none of this matters with respect to the current genocide of the Palestinians, but the modern politics overshadow the almost mundane aspect that I am more curious about regarding the movement and interactions of peoples from Eastern Europe to Central Asia prior to and after the Rus came into the picture.

I completely agree with genetics being useless.

Your point about 'most generalized' is unfortunately not accurate. Perhaps in the past, there were for example major issues with the reference human genome which overrepresented people from western eurasia, and only have a handful of ethnic groups from Africa (recall, any two random people chosen from Africa will be more genetically distinct than any two other humans; Papua New Guinea might be an exception and other parts of Micronesia; at least to the best of my knowledge). Since for example there were many people from Amerika also sequenced, it only added to the bias towards western eurasians.

The current reference human genome is very very much improved over the initial attempts. Not perfect of course, however more specific databases and other references which are specific to other regions/groups/etc (really whatever grouping of genomes may be of interest), one which is not really related to ethnicity is the neanderthal genome, the Human Pangenome Reference tries to take care of the issues with any kind of bias by having a collection of genomes. There's also the African Variation Reference Genome.

What does this mean? The resolution is increasing, note earlier that if only western eurasians and certain ethnic groups from Africa, like Mbuti, Yoruba, etc. if one was not of that population, say a Nuxalk Nation individual, enough would be different that it would be hard to say maybe if they were closer to western eurasian population A or B. Which says nothing of how they relate to the people living in close proximity, say Inuit to the northish and Anishinaabe to the east.

That isn't the case now. There are single point mutations, groupings of a few bits of DNA, haplotypes (seminal paper here), which one can look at and say with confidence, "This genome belongs to a person who is very likely from this geographic region." This doesn't mean people don't have variation, only, the genome is conserved, i.e. it doesn't change all that much especially in some areas of the genome which are like super important. Such as the genes related to DNA Replication, making tRNA; stuff where if it wasn't there the cell couldn't grow and replicate at a high enough rate for the organism to be viable.

You can imagine there can be a test, like a pH strip, which only checks for one of several characteristic haplotypes or combinations of haplotypes if additional resolution is needed, and the result would be quite accurate. Enough to discriminate at least. Enough to arrest and test more thoroughly, which is only getting cheaper by the day.

You need to accept that this stuff is getting better, and outside of medical use, mapping migratory paths, etc. there need not be any application in a way which is discriminatory. Denying the effectiveness is denying scientific progress, maybe not now, but at least in the future. If we are to be dialectical materialists then it must be considered that a compound with a given causal history may have the ability to inform those who study it of its history, and, that ability will change over time. The incoming nefarious uses can only be countered if they are accurately anticipated. Humans exist and have existed in a specific physical space at specific times, like layers of sediment or the wear on a rockface, it can be analyzed and information of interest can be determined with great accuracy.

Look at this database, this population, Ecuador Cayapa has a 45.8% chance of having the haplotype DPA101:03-DPB104:02. You might say, "The sample size is only 183" and you would be right. What you aren't including, is the hundreds of thousands sampled (this is only for high quality data which meets the needs of this database, there is much much more; typically the data needs to be de-identified and meet a laundry list of requirements to even be collected, let alone shared) which represent thousands of different populations spread out across the globe. And, though the haplotype I gave had 'only' ad 45.8% chance, and if you check it out other populations have it too, you'll notice on the page there are three haplotypes which have a very high frequency. That makes it very precise. Here is the associated global map, there aren't very many places where you would find an individual with the given haplotype.

This doesn't include all the haplotypes which this population likely or very likely does not have. With that it's rather trivial to determine where an individual is likely from, than to infer the migratory pathways and history.

Sorry for the long post, I just wanted to convey that I knew a bit about what I was talking about. Not too much, others who do this for a living or have advanced degrees would know much more.

For some good news, look at this database which has data on which haplotypes are associated with adverse drug reactions.

A note on variation. There is a genome, some areas vary a lot, some don't as much. There are other 'omes. Omics captures the entirety of the field of this kind of computational biology. Transcriptome is more or less a time-series analysis of the level of RNA transcript of interest at given points in time. Epigenome is the epigenetic stuff, so like, there are molecules/complexes, the usual example is a methylated site. A methyl group is attached at some point along the genome and it changes how a gene is exposed. That changes how often it gets transcribed, how often a gene is expressed. You can't necessarily look at the genome to determine which sites are methylated. I don't remember the inheritance, it was sorta complicated and it's been a while. mtDNA, or mitochondrial DNA can also vary, there are some diseases which lead to a non-viable fetus or a baby which will not survive past a certain age because their body doesn't like produce enough energy needed.

I made a response above, it is complex, and not that complex if enough subjects and enough data is attained. The more time goes on the more trivial it becomes.

The genetic component need not be the determiner for whether people of a faith, a culture, are discriminated against. Period. One need not react negatively to scientific claims, a good measure is, what if this limitation was overcome? What exactly would that change?

As far as I am concerned, with respect to genetics outside of medical use, or few other areas, nothing.