You might be right about that, although that is really terrible phrasing if that is the correct interpretation. Here's some more stuff from the actual article you linked as I'm going through it.
First of all, the article discusses a full sequence for one Korean male. Not a population, just ONE individual. In fact, the article is only using five distinct genomes in its entirety, and each one was sequenced from a single individual. Whatever you mentioned in your post above (here ) is either completely wrong, or this is not the study you were thinking of. This was not "a bunch of scientists" by any measure.
In the abstract, it mentions
This looks like they're saying they found significant differences even from closely related ethnic groups, primarily Korean and Chinese, the two groups they focused on in their study. They go on to say that constructing reference genomes for minor socio-ethnic groups would be useful. I take that to mean even small variations in socio-ethnic group membership would have enough genetic variation for separability or distinguishing between those groups. That might be a little over-reaching on my part though.
Later in the study it says
I'll help decode that a bit. SJK is the Korean genome they sequenced. YH is a single Chinese sequenced genome. NA18507 (Yoruba) is a sequenced genome from a Nigerian. Venter and Watson are individual genomes from two Caucasian men. SNPs are single nucleotide polymorphisms, effectively a single difference of a nucleotide. A difference in one SNP would be one base in the place of another at the same point in a similar genetic sequence, like a "C" instead of a "T".
The study is saying that the Korean genome shared 60% of these SNPs with the Chinese genome, 50% and 53% with Watson and Venter respectively, and 56% with the Yoruba genome, which is a Nigerian genome. The Chinese genome is 54% and 52% similar to Watson and Venter and 57% similar to the Nigerian genome. Watson and Venter are 56% similar to each other, and Venter is 53% similar to the Nigerian genome.
Naturally, this does not indicate which SNPs are similar in each case, just the ratio of similar and dissimilar between each group. I'm sure the individual SNP variations between the groups would be much more telling, if elaborated on. With a little bad science and hand waving, I can proclaim from these results that Koreans and Chinese are more similar to niggers than they are to whites, and more similar to niggers than whites are to niggers.
A more reasonable statement might be that Koreans are more genetically similar to Chinese than Whites are to each other, given Watson and Venter's only 56% similarity. The biggest question is the ethnic ancestry of Watson and Venter, because they could both come from very different areas of the world and may have significant admixture present. I'd have to read more on studies comparing the two of their genomes specifically, and a breakdown of their individual ethnic lineage. In addition, I wouldn't be surprised if there was more genetic diversity in Europe than Africa or Asia, which might explain the seemingly large variance between two Caucasians.
It's important to consider that this study is only referring to single genomes from single individuals though - it is absolutely not a population based study. The conclusions you could draw from this are marginal at best because there is no way to cluster four individuals from four distinct groups into anything but those four distinct groups. Even metrics of similarity are specious at best because they are single samples.