In answer to your last sentence Teristella - GOOD LORD YES!
Turning to your first sentence, the problem with performing an extended cross-match is that, if, for example, the patient has either sprouted a weak anti-Fya, or worse, has made an anti-Fya in the past (unrecognised), but has not been re-stimulated for a long time. Either the units could both be Fy(a+b+), or one could be Fy(a+b+) and the other Fy(a-b+), and the cross-match may not detect it. This is because the red cells in the antibody identification panel are in a preservative that maximises the expression of the red cell antigens, but NOT the oxygen carrying capacity of the red cells, whereas the red cells in the units are in a preservative that maximises the oxygen carrying capacity of the red cells, but does not maximise the antigen expression of the red cells (and the Duffy antigens are known for deteriorating upon storage (as are the Knops antigens, but they don't matter). So, by cross-matching without performing a panel, you may be missing an anti-Fya, and if the screening cell that is Fy(a+b-) may have come from an individual who has the FYA/FY genotype, if these red cells have come from a donor of the Black ethnicities, but the human immune system is MUCH more sensitive than our tests, so you have a DHTR on your hands.
This, however, is purely a personal point-of-view.