If a near-perfect repair capability were available, I can't see that it would be very costly to the body to, for instance, use it on stem cells, progenitor cell

If a near-perfect repair capability were available, I can't see that it would be very costly to the body to, for instance, use it on stem cells, progenitor cells, any -blast sufficiently up the hierarchy. I can't see it being all that energy intensive. It seems to me that DNA manipulation is relatively efficient, double-stranded break repairs via homologous recombination or NHEJ is in every cell. Bread and butter for your skin cells. And surely the energy consumption of all the turnover in epidermal cells alone would vastly overshadows the hypothetical cost of DNA repair in -blasts? We further spend serious amounts of energy on the immune system, on the brain, on all sorts of things, and we have good energy regulation mechanisms. Compared to all the things the body already does, spending energy on DNA repair seems like an excellent trade, when not actively underfed. I'm also suspicious of the argument that individuals living longer would be a detriment. DNA repair would not just cause longer life in the sense of very old people sticking around longer. It'd mean individuals living healthier lives for longer, so that they may still contribute to society, take care of grandkids, etc. Raising children is a big cost. If people lived longer healthy lives, arguably they could afford to have kids later if they chose to, which I imagine would save resources (e.g. by having a smaller percentage of life being spent in infancy and growth), potentially increasing fitness? A different argument still is that the body seems to have evolved many mechanisms to prevent cancer and to fight it when it happens. It seems that there has been selection pressure in the direction against DNA damage being a cause of death. If there were fitness in making people live shorter lives, I'd expect the body to try to do this not by letting DNA damage accumulate, but by trying to keep people as healthy as possible during their life, then separately using a programmed end-of-life to enforce a limit. I can't find the reason why this would be significantly hard for evolution to come up with, aside from the pristine DNA repair mechanism actually existing as described, which I don't think is really the case, no?