Joe says you can induce seedless fruit on pummelo by spraying with GA3 (Gibberellic Acid) when the plant is at 50% bloom. He wrote to not spray before bloom because that would prevent blooming. Before the bloom opens, there are reproductive buds, or unopened flower buds that are pyramidal shaped and about 1 to 2 mm diameter. If you spray GA3 at that stage, they will revert back into vegetative buds, thus you prevent blooming.
Technically, you minimize blooming when you spray GA3. Because of non-uniformity of statistical distribution of stages of bloom, a few of the flower buds would be at the point of no-return and wouldn't revert back to vegetative stage, but there is a stage of the unopened flower bud that can revert back into vegetative buds when sprayed with GA3. During 50% bloom, I meant that approximately 50% of the flowers have opened and are receptive for pollination.
We know that this is a population of distribution of stages of flowers, numbering in the hundred thousands or millions within a canopy, and for sure they don't have the same phenostage, and you would have various stages from unopened flower buds to petal fall, but approximately 50% of them are open and are receptive for pollination. When the flowers have opened and they are sprayed with GA3, the flower is tricked into reacting that it has been pollinated and so the fruit would set without forming any seeds, and thus we have seedless fruit.
This has been observed with watermelons, and other fruits. Again, it is not 100% seedlessness because of the population distribution of the stages of flower, and some of them have already been pollinated before the application of GA3 but majority of the fruits would be seedless, to those that were receptive but have not been pollinated and received a whiff of GA3, and then you would have tremendous fruit sets that are seedless.
Because you would have tremendous fruit set, you may need to spend extra labour to thin out the crop but that depends on your flower density, and so you may have to apply GA3 either at 25% open blooms or even at 10% but this depends on flower density. If you have few flowers, the 50% bloom would be the best time, but if you have a tremendous amount of flowers, you may need to spray during the 10% bloom. Some of the blooms would revert back, depending on the stage of the flower, but those that are open and are receptive for pollination would have fruit set that are seedless when sprayed with GA3. You can search for the references about GA3 application in citruses on how to control the alternate year bearing in Clementines.