I'm not sure why you think infertility rates of 3% (or more) work against the evolving of life forms. Why would they?
As we know, reproduction is not perfect. Mistakes and random reshufflings of genes occur, and the resultant offspring have some variation. Most of the mistakes are neutral or negative in effect on reproduction. Infertility would be an example of one of these. There's no way infertility gives a reproductive advantage. It's a mistake, but it's part of the process of evolution, rather than being an objection to it.
Not very often, variations ('mistakes') occur which
do give a small advantage to eventual reproduction. These can eventually spread throughout a population, because they give such an advantage, and so the DNA of the population gradually changes. Repeat that process over millions of years, countless small changes, and you can get speciation.
Infertility is an interesting question, in terms of evolution. It should be selected against, of course. Some simpler organisms have fertility rates of near enough 100%. More complex organisms, like ourselves, don't. Reproducing is an intricate process (surely
not intelligently designed); there's a lot to go wrong. Being born infertile is one of those tragic birth defects caused by the imperfection of the reproductive process - but it is just this imperfection in the process that helps to drive evolution, by occasionally throwing up a chance 'improvement'.
A disussion of evolution and infertility -
The evolution of infertility. (PDF)