I'm betting you'd rather say that "Mechanical twist actually twists the wing from root to tip to actually vary the angle of attack of the wing along that semispan."

Also, if you look closely at the elliptical wing and compare it to the rectangular one, you'll see that the rectangular one definitely shows the stall originating at the TE at the root of each halfspan. Compare the arrows that show how the stall migrates out from one specific location, the TE @ the root, to the similar function arrows showing the migration of the stall on the elliptical planform. That elliptical stall's arrows come from along a significantly wide span of the wing's TE. You are correct, elliptical wings do not stall at the root TE, however the picture doesn't show that happening.