In the comments of my first posting, J.T. pointed to a wonderful article precisely on point. I wanted to comment about the article. First, the authors do all the conventional research one can do to prove a relationship. In fact, if Gilbert Lane had used his daughter's married name instead of just her given name of Eleanor, there would have been no need to do the DNA testing. So, the researcher was right there in the right place. As to my thoughts, this article gives large credence to how hard (and long) the process would be. In this article, the research problem is a good 50 years younger than mine (ca. 1800) and because it lies on the maternal side of the researcher, he had one DNA match in his great-aunt already. So, he started off on second base for his home run. Still, to get the other half of the DNA proof, he runs into all the problems I mentioned including a match who would not cooperate. [That person had "issues."] Still and all, they spent the time and money and proved things. I would have liked a more detailed proof at the end, however. One still must say that based on the research done and the DNA analysis from the two candidates that this is the most likely solution. There is still a chance that the Eleanor in question is not a Lane, but someone else who just happens to be in the same mitochondrial line. How likely is that? Well, how many female to female descendants of Annetje Cool are living in New Jersey and New York in 1800? Tough call. That is always my problem with DNA. It just means that "at some point in time" the two lines are connected. When in time is always a question.
Most of the Y DNA studies I've seen say that line X must be from ancestor A, but leave several generations blank in between because of the lack of a paper trail. Still, this is probably the beginning of the future of genealogy. A century from now, future scholars will re-write some of our work based on DNA, just as we rewrite past scholars work in light of newly found primary evidence.