I'm not going to be tremendously helpful, Barb, but some thoughts do occur spring-boarding off Susan's and Tom's comments. First, I don't know of any uses of 20 years as an average generational interval other than anthropological studies trying to reach as far back as the Neolithic. For genealogical relevance, the average generation length is longer.
For estimates, I typically use 32 years for males and 26 years for females--sex-averaged at 29--based primarily on work by anthropologist Jack Fenner (Jack N. Fenner, "Cross-cultural estimation of the human generation interval for use in genetics-based population divergence studies," American Journal of Physical Anthropology 128 (October 2005): 415-423. https://doi.org/10.1002/ajpa.20188) who found average generation intervals in hunter-gatherer societies of 31.5 years for males and 25.6 years for females; 30.8 and 27.3 years, respectively, in males and females in more developed countries...so still 29 years if sex-averaged. That's in pretty close agreement to other research like Helgason et.al. (American Journal of Genetics, 2003) 31.9/28.7 [male/female values]; Matsumura and Forster (Proceedings of the Royal Society of Biological Sciences, 2008) 32/27; Donn Devine (Ancestry Magazine, 2005) 33/29. Turi King and Mark Jobling (Molecular Biology and Evolution, 2009) settled on 35 years for males of British origin.
Nit-picking, I know. But the difference adds up quickly if working TMRCA for many generations back, and the male/female distinction becomes important when dealing with yDNA. Using Turi King's estimate of 35 years (from the 2009 publication mentioned above and another paper in Current Biology, 2006), a 15-generation TMRCA would be estimated at 525 years before the birth of the current DNA test-taker, so call it about 1445 AD, shortly after the time when surname adoption was becoming common in the British Isles. The sex-averaged 29 year interval at 15 generations would put us at about 1535.
I for one am guilty of frequent use of the term "genealogical timeframe"...something which doesn't have a firmly established definition. Family Tree DNA defines it as, "The genealogical time frame is the most recent one to fifteen generations." Wikipedia approximates it as "c. 1500 onwards." ISOGG has one of the most liberal definitions: "the period in which it is possible to find genealogical records relating to individual ancestors." The BCG has no specific definition. As common usage, though, it is the period in which, for a particular region, surnames were first consistently adopted as an identification of lineage.
The reason is simple: evaluating genealogical information is extraordinarily difficult if there is no consistent naming record from generation to generation. Standardized naming systems are probably oldest in China; in the time of the Shang Dynasty surnames had become patrilineal. Confucious's family tree is supposedly 80 generations in depth. But in the 5th century, the use of family names was unusual in the Eastern Roman Empire, and they weren't used at all in Western Europe when the empire fell. The practice started gaining popularity in the 9th or 10th century around the Caucasus in Eastern Europe, and began to spread west. The first known use of a surname in the British Isles was in Ireland, in 916. It was the Domesday Book in 1086 and the advent of taxation by the Normans that led England to start to see consistent surnames, but it was still rare into the 13th century. It would be another century or two before the practice had gained popular ground in the British Isles, and some Scandinavian countries didn't move away from patronymics until the 18th century. It wasn't until the early 19th century, and a decree by Napoleon, that the then-French-annexed Netherlands had to adopt and register distinct surnames.
So Tom's point that "any [15th generation ancestor] would have so little bearing on heredity as to be meaningless" is well-taken. I'm as much or more geneticist as genealogist, and I admit I blank-out and begin thinking about what bills are due or what to make for dinner when the topic is a family tree that's, say, 25 generations old. In the case of Western European descent it's almost impossible to prove, and it makes so much more sense to me to pay most attention to, as Susan points out, the accuracy and breadth of the most recent 10 or so generations. I don't know why WikiTree decided to make the Family List and Dynamic Tree go only to 15 generations, but since that should put us back about the mid-1500s I think it's a reasonable demarcation point.
As Susan said, though, adding the "improvements" tag to the question would bring it to the attention of the folks who could look at possibly extending the reports beyond 15 generations.
