I happened across this paper that points out a the low incidence of cancer in countries where kava is consumed (pdf attached). There are a lot of problems with the paper (the apples-to-oranges comparison of reported incidence of disease in developed vs. developing countries; the uncertainty of trying to assign causality to one particular aspect of a culture to a specific statistic, etc.). It's not a rigorous paper at all, but it is interesting to consider this in light of recent finding laboratory findings that kava (and flavokavains in particular) may have a protective effect against cancer.
http://www.sciencedirect.com/science/article/pii/S0944711316302045
But the level of rigor is comparable to the studies showing elevated GGT enzymes in small cohorts, so I'm throwing it out there for discussion. Any speculation about the effect kava consumption might have on overall population mortality levels (and there is no evidence that it has any effect, either pro or con), needs to consider this.
Note the above in vitro study has statistics for the specific case of colon cancer in the Pacific Islands, but also notes that they are low. Quote:
"There is a low incidence of colon cancer in the South Pacific Islands, including Fiji, West Samoa, and Vanuatu (Steiner, 2000; Foliaki. et al., 2011). The incidence of cancer (2000–2005) is lower for people in the Pacific (Tonga, Fiji, Cook Islands, Niue) compared to Pacific people living in New Zealand (Age-standardized cancer incidence rates: 315 per 100,00 person-years in females, 379 in males), which is similar to the rates for New Zealand overall. For Fiji, the rates were 231 and 126 (per 100,00 person-years for females and males), respectively."
It is especially interesting to note the gender difference in cancer incidence, since in traditional kava drinking societies (and Fiji might be the best example today where the traditional culture still exists to a large extent) kava is only drunk by men.