Genetic Load
The reduction in fitness in a population is proportional to the deleterious mutation rate (Haldane 1937). The genetic load for a single gene is
[math]\bar{w}=1-\mu[/math]
So, if the reduction in fitness is independent among genes and multaplicative for n genes it is
[math]\bar{w}=\left(1-\mu\right)^n \approx e^{-n\mu}[/math].
Define the genomic deleterious mutation rate, over all n basepairs in the genome, as
[math]U = n\mu[/math]
then
[math]\bar{w} \approx e^{-U}[/math].
If we have three to eight deleterious mutations per genome per generation then
[math]\bar{w} \approx e^{-U} = e^{-3} \approx 0.04979[/math]
[math]\bar{w} \approx e^{-U} = e^{-8} \approx 0.00033546 [/math].
Being able to generate a wide range of complexity from a small number of underlying genes may be a way to evolution to deal with some of the problems of genetic load. There are less genetic targets for mutations but the phenotype is very flexible and evolvable (versus many genes to fine tune various details of a phenotype).
http://blog.wolfram.com/2009/03/25/minimum-inventory-maximum-diversity/
https://www.nature.com/news/2003/030402/full/news030331-3.html
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0029324
https://phys.org/news/2018-09-chaos-inducing-genetic-approach-stymies-antibiotic-resistant.html