Someone needs to mention an important fact the OP might not have considered. Most self injurers do not kill themselves from the injury process or even intend to. They end up with a lot of scars and and bruises, but for the most part they survive. How would behaviors that rarely cause severe/irreparable damage in the long run be a means of removing yourself from the gene pool?
In regards to
suicide, at least in the US, old people who cannot have kids are some of the most successful when they attempt it and people in their forties and fifties (who have probably had kids already) have the most successful suicides of any age range. Teens, children, and young adults who have the most reproductive potential have the lowest successful suicide rate of any age group. Even the total number of young people who successfully complete their suicides is only higher than old people who make up a significantly smaller (over 5 million fewer people) part of the population. If suicide is supposed to remove people from the gene pool, it is not very good about ensuring success at the right time. This is aside from the logical fallacies involved in the theory and its misconceptions about evolution, genetics, and human behavior.