"Since evolution cannot foresee the distant future, in which sexual reproduction proves to be superior, how did sexual reproduction ever get started?"
One likely scenario would involve four basic steps. The first would be to change a unicellular organism from haploid to diploid. That is, instead of carrying only one copy of each gene, it now starts to carry two copies of each gene. These additional copies, called alleles, would actually accelerate evolution, because now one allele can be changed into a different gene without endangering the organism. Conjugation would allow different members of a population with different alleles to share them with other members of the population.
The next step could not occur until after the unicellular organism had evolved into a multicellular organism. At that point certain cells could mutate so that they undergo meiosis instead of mitosis. Being part of an integrated group of cells would allow them to make this change without killing them off. The next step would then involve packaging the nuclear material of the meiotic cells, perhaps as a way to try to repair the "damaged" cells. Finally the last step would involve expelling this packaged meiotic nuclear material to get rid of it. Once outside the organism, however, these packages could merge and form new viable cells. Once again evolution would accelerate, as different combinations of alleles would recombine to create a vast number of widely unique individuals, capable of surviving in a wide range of habitat niches.
Beyond these four basic steps, natural selection would then fine tune the sexual process to make it as efficient as possible.
"Who addresses this problem? No one, to my knowledge."
This is not true. A search of any electronic biomedical literature database using the keywords "sex" and "evolution" will bring up dozens of articles in the last five years alone. There is in fact a fair amount of research being done on this subject.
Kevin L. O'Brien