I agree Servals and house cats share many of the same genes, and, thus, most genes are homologous and their alleles are also mostly shared. And, since Servals have, to the best of my knowledge, never been sequenced, we have no idea which are homologous, which alleles of homologous genes are shared, and which genes are unique to each species.
All we can really say at this point is that a F1 from a Serval and a house cat will have 19 chromosomes from the Serval and 19 from the house cat. That's the 50% I'm referring to.
If you refer to percentage of genes, then most F1 males will have fewer genes from the Serval father (obviously for most F1s the serval is the father) because the inherited Y chromosome is shorter and has less genes in it than the X chromosome inherited from the house cat mother. However, as I understand it, the expression of genes on the Y chromosome are not mitigated for the most part by genes on the X chromosome and thus are mostly expressed, so a male F1 may have a phenotype that is more "Servally".
Once you go beyond F1, or with a HP F1, you can no longer say with any certainty what the percentage is because of random chromosomal crossover.
Of course, now we understand that there is much more to genetics than chromosomes and associated DNA. There's mitochondrial DNA, which is only inherited from the mother, making most F1s even less Serval. Eggs and sperm also carry RNA--but less in sperm. This RNA, epigenetics, environment, dominance of genes in all of it's complexity, and other influences affect gene expression. So a F1 could have many Serval genes that aren't expressed, but could still be inherited by offspring.
As we can't calculate disparate genetics, I'll stick to 50% of chromosomes for this particular F1 mix is all we can definitively say.
Ziggy's breeder is a geneticist and we've had a few discussions on this topic. However, I'm an amateur at best and all mistakes are mine.