That's an interesting observation. Do you have any literature you can point me too, that backs that up? I read that UT is best suited for detecting planar flaws, while x-ray and UT are equally suitable for detecting volumetric flaws (see below).
I suppose the thinking is that, you could also have planar flaws. Thus, UT is sensible because it could pick up flaws that X-ray couldn't.
It seems though that 100% inspection is overkill in certain cases based on part dimensions and weld characteristics.
For reference, the paper with the comparison between x-ray and UT is:
Jack Spanner, “Effectiveness of Substituting Ultrasonic
Testing for Radiographic Testing for Repair/Replacement”
ASME Conf. Proc. 2005, 33 (2005)