Re: I have a question about the percentage of alleles that cheetahs share.

Yes, you are correct.  The figure they are talking about is for alleles, 
not genes, and they are talking about the most unrelated individuals 
considered to be of the same species, not the most related (identical 
twins of course share 100%, siblings average 50% or more depending on how 
related the parents were, etc. and depending on how similar two alleles 
have to be before they are considered "identical alleles").

The genetic similarity between all cheetahs currently alive in the world 
indicates that they passed through a severe genetic "bottleneck", in 
relatively recent times such that diverse alleles have not had time to 
arrise de novo since the bottleneck event.  Humans also seem to have 
passed through a fairly narrow bottleneck tens of thousands of years ago 
(or more) but those tens of thousands of years have been enough time for 
us to create some genetic diversity since then.

There is a huge difference between sharing some percentage of identical 
alleles, and percentage of overall genetic similarity.  If two alleles of 
a gene that is 10,000 DNA bases long have a single point mutation, they 
may or may not be considered "different alleles" depending on how the 
allelic difference is measured.  Some alleles are measured as different 
only if they result in a detectably different phenotye.  Other alleles 
are measured as different only by DNA sequencing or other biochemical 
means, that would cause a visibly different phenotype.  At any rate, if 
the single base change in the 10,000 base gene did result in enough of a 
change to be considered a different allele (for example, single base 
changes in the hemoglobin gene can result is a sickle cell phenotype and 
be counted as seperate alleles) then the allele is counted as "different" 
(as if it had zero similarity) when the genetic similarity between the 
two alleles is 99.99% or one base differnece in 10,0000.

Percentage of DNA sequence similarity is a very exact number that can be 
determined by sequencing very large regions of the genome of two 
individual members of a species or between species.  Percentage of shared 
alleles is a "fuzzier" number that can vary depending upon what methods 
are used to compute differences in alleles.  Some types of organisms 
generate huge variations in visible phenotypes with vary little genetic 
diversity (for example, all domestic dogs shared a very small genetic 
bottleneck within the past 50,000 years) while others all look visibly 
the same despite huge genetic diversity (for example all earthworms, even 
from different species, look the same to an untrained eye).

It is common for news writers, even "scientific" news writers, 
to write meaningless or outright incorrect statements about scientific 
subjects such as chemistry, genetics or computer science.  We really do 
not have a clue how much allelic diversity is present in "most species".  
For one thing, we can't even define "species" for most organisms such as 
bacteria, viruses, protists and many fungi and plants.  For another 
thing, of those organisms where the term species can be clearly defined 
(such as for most mammals) we have not done careful sampling and allele 
diversity analysis of the most diverse members of the species.

A more correct statement might have been to compare allelic diversity in 
cheetahs to allelic diversity, measured in the same way, in lions, 
cougars, leopards and several other well-studied members of the felines.  
Measurements of allelic diversity can be very important for determining 
many factors that influence the ecological "health" of a species, such as 
populations sizes and ability to make contact between isolated 
populations.  For example birds are less influenced by habitat 
fragmentation (for example building a lake or road or city in between to 
breeding populations) than are many other animals.  
  
The scientific field of ecological genetics is booming now, because 
humans are causing the most massive extinction of species (especially 
vertebrates) seen on earth in the past few million years.  If we have to 
choose which species to save and which individuals of a species, we can 
make more intelligent desicions if we know something about genetics.  A 
few species live only in captivity, in zoos or animal parks, and it is 
important to use breeding programs that preserve as much genetic 
diversity of the remaining individuals as we can.  Other species are in 
danger because of habitat fragmentation, and we can seek ways to minimize 
this by intelligent urban (and rural) planning.