In many ways I do agree with what you write, but, however rare K+k- individuals may be, they are considerably more common than are either Kp(b-) and Js(b-) individuals. It would seem reasonable, therefore, to test for such individuals, in the knowledge that, nationally, and, perhaps, internationally, K+k- blood can be supplied reasonably easily, whereas the search for a reasonable supply of either Kp(b-) or Js(b-) donations could prove fruitless.
Again, with regard to Dombrock antibodies, although undeniably these can be clinically significant, such clinically significant antibodies are incredibly rare - so much so that they are often still written up as posters or abstracts, even in patients who are not on Dara, or similar monoclonal antibody treatment.