Sigh. Somehow, I thought we had beat this poor horse senseless already. It is physically impossible for the mtDNA X haplogroups listed above to even match each other, and none of them match the mtDNA of the Anzick Child.
On the X haplogroup: Reidla, Kivisild, Metspalu, et. al. "Origin and Diffusion of mtDNA Haplogroup X," American Journal of Human Genetics (Nov 2003) 73(5): 1178–1190; doi: 10.1086/379380. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1180497/.
An X2a sample cannot match X2b or X2c samples, even at the low-resolution HVR1/HVR2 testing levels. They are distinct branches of the X haplogroup phylogenetic tree (click here to view image of tree), and are defined by mutations that don't coexist with each other.
Haplogroup X is not Native American. It bifurcated from Haplogroup N roughly 30,000ybp. It's found in a broad geographical area, including Europe, the Near East, Central Asia, North Africa, and North America. Haplogroup X2a does seem to be a uniquely American First Peoples subclade, and may have migrated to the Americas roughly 15,000 or so years ago.
What data exists so far regarding X2a seems to break down like this (showing the mutational variances in the last three columns):
||Native American/First Peoples
||8913G, 12397G, 14502C|
||North American Great Lakes/Plains
||Same as X2a1
||Same as X2a1
||Same as X2a1
||Nuu-Chah-Nulth, Na-Dene-Navaho, Yakima
I've never seen any credible evidence that the data from Anzick-1's sample was ever tampered with or modified. And the Anzick Child isn't even of the X haplogroup; Anzick-1's mtDNA is D4h3a (Rasmussen, Anzick, Morten, et al. "The Genome of a Late Pleistocene Human from a Clovis Burial Site in Western Montana," Nature (Feb 2014) 506: 225–229; doi: 10.1038/nature13025; PMC 4878442. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4878442/), and you can view the D4h tree here. The D4h3a designation was confirmed by two different types of tests: traditional Sanger sequencing as well as whole-genome shotgun sequencing.
Haplogroup D bifurcated from haplogroup M sometime around 40,800ybp, probably in Asia. "The subclade D4 is the most frequently occurring mtDNA haplogroup among modern populations of northern East Asia, such as Japanese, Okinawans, Koreans, and Mongolic- or Tungusic-speaking populations of northern China." (Wikipedia) The subclade D4h3 is found most commonly in the area of Thailand and Laos; D4h3a in Peru, Ecuador, Argentina, Bolivia, Brazil, Mexico, and the US.
All mtDNA haplogroups considered indigenous outside of Africa descend from either N (the parent of haplo X) or M (the parent of haplo D). M is believed to have split from L3 over 60,000ybp, and N split from L3 over 71,000ybp. So the X2a clade found among American First Peoples is in no way related to the D4h Anzick Child. The two wouldn't have shared a matrilineal ancestor not just back to the Neolithic or even Upper Paleolithic; it would have been at the end of the Middle Paleolithic. Maybe around or somewhat after the Toba supereruption that's been hypothesized to have created a huge human population bottleneck.
The Kennewick Man is, as Loretta noted, X2a...in fact, his remains represent the most basal X2a lineage yet discovered. Speaking of, the article that started this particular thread is by Jennifer Raff and Deborah Bolnick (you can read it here). A final note is that Jennifer is a great follow on Twitter, if you're so inclined. Not just genetics and anthropology; she's a pretty eclectic Tweet-maestro. Find her at https://twitter.com/JenniferRaff.