Hi Alan, my sister has been researching our Stewarts for many decades and uses Ancestry and LDS records.
I decided to try WikiTree for my work (I dont use ancestry.The information on WikiTree re William born 1770 Clunie and his descendants has mostly been done by her, with some additions by the person in charge of the Stewart project on Wikitree.
I have concentrated on verifying the information through OPRs via Scotlands People and trying really really hard to find this darned James via Scotlands People, My heritage, LDS records - which my sister used for an enormous amount of research over decades, the Red Book and old documents such as Commissariat reports, census etc. (I dont use ancestry). Her approach has been very linear, whereas I tend to also look for connections through branches. Your suggestion re tracking all the James's and their kids is great and I will start compiling those, thank you!
Re the DNA, we're neither very sure how that works, although I have watched the recent Stewart project video. My assumption is that using autosomal DNA, finding matches laterally works really well, finding matches in ancestors however must depend on someone with a verified line to that ancestor. It isn't until we come down 3 generations to Peter Stewart (following William born 1817) that we get a match, and that is our second cousin related through Peter Stewart's mother's side. When we come down one more generation to my grandfather who died at Somme, we have a match wth a cousin through Robertsons.
Am I right in assuming that the DNA match with William born 1770 is a probability, based on our assumptions that William 1770 is the father of William 1817? (if that were to be wrong I dont know what we would do, frankly.) This is both addictive and incredibly frustrating