They address this -- there's a discussion of the likelihood of laboratory evolution, and their conclusion is that it's unlikely because the progenitor virus has

They address this -- there's a discussion of the likelihood of laboratory evolution, and their conclusion is that it's unlikely because the progenitor virus has never been described, and the selection system has never been described. I don't think this is the strongest argument they could have made. To me, the strongest counterargument is that the large insertion (4 amino acids; 12 nucleotides) is statistically unlikely. Even highly efficient laboratory selection systems don't produce those kinds of mutations very often. Maybe you could do it by passing the virus between pangolins or bats over and over again, but that would be an immense amount of work. Is it possible? Sure. Likely? No. The parsimonious answer is that it's been circulating in nature.