When those pistillate hermaphrodite fig that produce the occasional viable seed self, you get a genetic copy of ga/ga female as the plant selfed itself. You are not going to get any males, no San Pedro and no Smyrna fig! You get all genetic copies of the mother! So why when you take ga/ga 'feminized' pollen and use it in a cross on a different ga/ga edible fig (not Desert King) and get 1/3, 1/3, 1/3??? It does not make sense!
When making a synthetic S1 cross using 'feminized' ga/ga pollen, I EXPECT 100% female common edible fig all of pretty good quality, just a blending of the two superior female phenos ga/ga.....The only minor differences should be the minor amount of soma clonal variation that takes place during genetic recombination during meiosis. However there is very little crossing over during meiosis in plants and it usually just results in differences in yield. But his being said, the Japanese are using this simple technique for a reason...IT WORKS!!! If it works....ALL FIG ARE HERMAPHRODITE! Go back to this:
Staminate male(GA/ga..GA/GA??)-----> Staminate hermie(GA/ga...GA/GA??)------>Hermie(GA/ga)<-----Pistillate hermie(ga/ga)<-------Pistillate(ga/ga)
If you can ignore 3 out of the 5 and use pistillate hermie ga/ga and pistillate ga/ga in synthetic crosses using feminized ga/ga pollen....why do anything else? Unless were do not have enough landrace fig to work with and that appears not to be the case!