I've been looking at the Keziah Arroyah story to get a better grasp of the situation, and they're right, the timeline doesn't work at all. It's possible that there's more than one version of the story, but the one I've seen in multiple places is that Ka Okee married Thomas Pettus around 1635 and gave birth to two daughters, Christian and Unknown. Ka-Okee evidently died sometime during 1638-1643, since it's documented that Thomas Pettus had married someone else by 1643.
It's documented that Christian (born 1636) married John Martin and then Francis Waddington. It's reported that Unknown (apparently born after 1636) married Chief Wahanganoche and gave birth to Keziah Arroyah. Given the limitations of nature, it doesn't look like Keziah could have been born any earlier than 1650.
Keziah reportedly married a Richard Bryant (1615-1680) whose existence has not actually been documented, and gave birth to the documented Richard Bryant (1651-1704) when according to the timeline she was not more than one year old, or possibly hadn't even been born yet. Ann Alleged-Meese was the wife of this second Richard Bryant. I was confused earlier, and thought that Keziah and Ann were thought to be wives of the same Richard Bryant.
I've seen a line of descent for Ann Alleged-Meese in which Ka-Okee herself marries Chief Wahanganoche, gives him a daughter Mary Ann around 1635 (which is the same time she supposedly married Thomas Pettus). Then Mary Wahanganoche marries Henry Meese, gives birth to Ann Meese around 1656, and Ann marries the documented Richard Bryant. This at least fits the timeline reasonably well, But as Kathryn Forbes pointed out, Ann's parents are unknown. Henry Meese can't be her father unless she was an unacknowledged illegitimate child. It does look like he had an off and on presence in Virginia around the time she was born, so its not impossible.
But I don't use "It's not impossible" as the standard for accepting the validity of a line of descent. It's not impossible that Ka-Okee was a real person who married Thomas Pettus and that Christian is their daughter. But there is no convincing evidence that it's true. It IS impossible that Keziah is the mother of the documented Richard Bryant. It's not impossible that the DNA project has identified a genuine family connection between the Christian Martin and Richard Bryant families, but all the talk about endogamy and pile-up regions makes it sound more like they've identified common ethnicity rather than a true common ancestor.
If there is a true common ancestor, it is NOT reasonable to declare that the she must have been the daughter of Pocahontas. It's not impossible that it was Ka-Okee, but really it could have been any woman in the region who married any random Englishman and had at least two children with him. The Pocahontas story is an interesting hypothesis that can't be ruled out, but it doesn't prove anything.