It's easy for a variant to become dominant just by random chance if the prevalence is low. That's the case here. Alpha and Delta were quite different, they took

It's easy for a variant to become dominant just by random chance if the prevalence is low. That's the case here. Alpha and Delta were quite different, they took over when the case counts were high, so it was basically impossible for founder effects to be the reason. So in this case we need to rely much more on the circumstantial evidence (which admittedly seems pretty solid on a first skim).