Not all genetic changes confer such a tiny percentage of a trait, though some might. Many, such as the gene for brown or blue eyes for example, act in an almost binary manner, with one trait dominant, but the other instantly at full expression when conferred by both mother and father. Others merely push the average of some quality, such as, say, frontal lobe size, or hormone density, in one direction or another, and that is enough to make observable, meaningful differences. Either way, even in your speculative scenario, genetic and phenotypic/behavioral/qualitative change does indeed take place, which is all one needs for selective forces to act upon. Thus, evolution must take place. You'd need one Hell of a lot of extraordinary proof to seriously suggest that it doesn't.