The excerpts are from Klimov's Etymological dictionary of Kartvelian languages.

 The excerpts are from Klimov's Etymological dictionary of Kartvelian languages. The proposed age of Proto-Kartvelian divergence is 4200 years before present. Others proposed slightly higher ages. Anyway it's safe to assume that it occured in Middle Bronze Age.

The first separation are Svans. Later Zan languages separated from Georgian ( or Kartli ) at 2600 years ago. While in medieval period Zan itself splits to Mingrelian and Laz. 

Georgian was the language spoken in historic Iberia. Which in medieval period expanded thanks to the emergence of unified Georgian kingdom under the rule of Bagratids. 

With current genetic and archaeological data combined the most likely place of proto Kartvelian divergence might be in Western Georgia in Bronze Age. Where they were dwelling before Bronze Age is a different subject. But in most likelihood their ultimate origin is in Fertile Crescent.

Gamkrelidze proposed a slightly Eastern homeland but current genetic data do not support that. First we can't place their divergence homeland in Eastern Georgia because in Western Georgia there is no Steppe ancestry. While Eastern Georgia might have a lot of it in MBA. If a migration from East to West occured in MBA that would definitely had left traces.

The other reason is the archaeology. Eastern Georgia had Kur-Araxian culture in EBA and Trialeti-Vanadzor in MBA. Non of this culture fits for Proto Kartvelian homeland.

Some believe that proto Kartvelian homeland was more southern toward Anatolia. But that's also is not feasible. Given that modern Western Georgians have excessive affinity to Eneolithic western Caucasus (CHG rich ) and less to Anatolia. Should an important migration from South to North have occured they would have more Anatolian ancestry than they have today. Which is not the case. Also their popular Y dna do not fit this theory.

So we are left with the only one solution. To place them in Western Georgia. 

Nevertheless this theory is not without problems. Because in Western Georgia there is also evidence of ancient NWC like onomastic. Gamkrelidze mentions evidence of NWC like substrate in Svans. Also there are very few hydronymes with obvious Kartvelian etymology. One of such a Kartvelian name according Klimov is Chorokh which appears in Greek sources much later than the Akampsis/Boas names. 

One solution to overcome those contradictions is to imagine that NWC related groups were initially more prevalent in coastal regions while Kartvelian were in mountain regions of Western Georgia. Later they expanded toward coastal regions. It is probably during those expansions that the term Sanni ( or Tsanni ) appears in Greek texts.