There may be a lot of differences on a micro level. "所", for example, is só͘ in Coanciu (lit.) but sé in Ciangciu (lit.). Coanciu literary readings are dominant in TW, and I didn't know the other reading existed till Elmer brought us this project. (This does explain why "初" is che, though.) If I had to bet, I would bet on Kakbe patterning with Ciangciu. But I wouldn't consider this a safe bet.
Coanciu literary readings being dominant in TW -- I didn't know this till now. It means I'm much less qualified for this than I had thought; I thought I knew the Ciangciu literary readings, but it turns out I don't.
The colloquial is its own animal. Kakbe is right up against "Tang'oann" territory. There may be similarities in vocab., etc. People had to communicate.
Back to the matter at hand, socio-linguistics and naming practices actually cloud the matter. Elmer's clan used a number of what could be considered "obscure characters", but a lot of these characters are just a common character plus a radical such as 王土金水艸 etc. U could imagine the family -- most likely merchants, farmers, or both -- consulting with a necromancer who tells them, "This child will need 土 in his life, so insert a 土 into his name," or "If U put a 金 in this child's name, the family will prosper." It is likely that the clan than used the reading of the "original" character, possibly w/o even knowing the obscure character that was then created. "俶", for example. My guess is that they just wanted "叔" with cheese. But it's hard to say. "瑱" is even more likely to be so. They wanted to use "真", but they wanted to "jade" their kid.
The people most likely to be able to "get this right" would be the "resident scholars" of the family left over in situ -- if the family still exists back there, and has scholars still living. But I think we can get it down with few discrepancies, if any (not really errors, just discrepancies).
