O being the original blood type hasn't been tested scientifically either - though it can't really be tested scientifically - a prejudice of the times perhaps - not dissimilar from ones of the past.
But if O is less opportunistic for diseases, that can be tested, given the will, and indeed it is a better explanation than the out of Africa argument combined with the distribution of blood type in the region. I guess we'll have to wait and see as to whether A and/or B provide more opportunities for disease than O. A combination of disease, civilization and migration seems like a pretty good bet to me. Africa might still come up as the place of origination, but then again it might not. But the testable hypothesis is sound.