Thanks for getting this back on track. I just want to make a quick note about percentages in genetics. While you can pretty confidently say the first generation is half of parent A and half of Parent B, you can’t say with any certainty what the percentages of subsequent generations will be. Why? Because the plant carries 2 copies of each gene, one from each parent (or in hexaploid persimmons, 6 copies, 3 from each parent). Now, which copy gets sent along to the next generation in a given ovule or pollen is determined at random for each gene. Basically, each gene is a coin toss whether parent A’s copy or parent B’s copy goes into the next generation. And while you’ll probably get half heads and half tails, it’s entirely possible to get all heads or all tails. So there is a vanishingly small chance two 50/50 hybrids bred together could produce offspring that have 100% kaki genes or 100% virginiana genes. What’s happens in practice is that the highest number of seedlings will be 50/50, with decreasing proportions on each side of that. Think of the classic bell curve. Now, if you breed a 50/50 with a 100% kaki, your distribution will be centered on the 75% kaki, but any given seedling could be anywhere on a broad spectrum from 100% kaki to 50% virginiana genes.
In sum, the fractional heritage is useful for predicting the genetics of a population of seedlings, but wouldn’t tell you about the genetics of any given seedling.