ian, I think that is a really interesting summary. Have you read "The sports gene" by David Epstein?
I haven't but I might. I suspect he didn't have Usain Bolt wading through treacle though, but if he's open to the experiment, I'm willing to put up a bag of hot chips...
I should also say, it's not genes
per se, it's more often a case of how they're regulated. It's rare that a mutation in a single protein has a significant effect, those early days of thinking one gene = one thing, are mostly gone (though everyone likes the simple elegance of Mendelian inheritance). Traits (and dieseases) are complex polygenic things, but more importantly, our phenotype is the result of a massively sophisticated system that starts with DNA, which is filled with regulatory sequences (there's far more regulatory DNA than coding DNA), the DNA generate the RNAs that make proteins, but also lots of small RNAs that regulate that process, and then that is all turned into protein. Those proteins may be structural, enzymic, and in turn, may regulate, modify, and package DNA (and RNA and other proteins). Then there's biochemical, endocrine and physiological regulation at all levels from organelle to entire body.
That's why, despite being near identical to a chimpanzee at the genetic level, humans generally aren't mistaken for chimpanzees.
Anyway, it's complicated, so that's why you should furrow your brows when someone says they have the gene or genes for so-and-so. Once you get beyond Mendel's wrinkled peas it rapidly complicates. Even most 'cancer genes' only deal in probabilities. Probably the most famous, BRCA1, comes with a significant chance of breast cancer in women, the risk for carriers of the mutant cancer-related variants is something like 80% over 90 years (and around 50% for ovarian cancer). But some people with the variant won't get cancer. BRCA1, on the other hand, only causes modest cancer risk in men. There are lots of other things happening (the BRCA1 gene encodes a protein involving in fixing particular types of DNA damage, the mutations result in low levels (a total absence of BRCA1 is generally fatal at the embryonic stage), but that damage has to occur first, and then the failure to fix it has to become significant (and we have all kinds of DNA repair pathways, your DNA is being broken, damaged, and inaccurately copied all the time). Which is why it comes down to probability, damage accumulates over time.
Incidentally, Mary-Claire King, in whose lab BRCA1 (and the role of its variants in causing cancer) was uncovered in, spent her early career scratching her head and wondering why she couldn't, at the protein level, find significant differences between humans and chimps (initially she assumed she was doing the experiments wrong). Of course, that lack of difference goes all the way down to DNA (but the technology to easily sequence entire genomes had yet to come along). But we're pretty similar to onions for that matter.