I'd like to see the materials that purport to demonstrate this proposition.
It is true that the amount of crimp clamping force that can be applied by a true taper crimp prior to the point of case deformation is less than the amount of crimp clamping force that can be obtained via a roll crimp, the notion that passing a loaded straight wall round through a taper crimp die (set otherwise than for zero crimp) results in zero crimp clamping force defies the laws of physics.
I meant excessive crimp, not a die set "otherwise than for zero crimp". A mistake that many newbies make when a bullet is not held tightly is to increase the taper crimp in an effort to improve the case's ability to hold the bullet. It's not the way to fix the problem.
Many die sets have a crimping die that taper crimps to a point, then roll crimps when adjusted further down. If you adjust this die down to "get a stronger crimp", it will begin to roll crimp. You'll deform the brass when the case mouth hits the lip inside the die that performs the roll crimp. Since the bullet has no cannelure, there's no place for the case mouth to roll into and the brass is pushed down. The brass immediately below the crimp will bulge out, away from the bullet. In this case, excessive crimp results in a net crimping force of less than zero when compared to an uncrimped round.
When using a true taper crimp die (where the inside of the die is only tapered, and the mouth of the case does not bottom out on a lip), excessive crimping force does not increase the "holding power" of the case because the metals involved, lead and brass, have different moduli of elasticity. (I apologize in advance if this is basic stuff - I don't mean to be pedantic). Try this, take a strip of lead and a strip of brass and put the ends in a vise. Take the ends that are sticking out and bend them the same amount. If you bend them a tiny bit, they'll spring back to their original positions. Bend them a little more and the brass will spring back all the way, but the lead will only partially return. You've exceeded to modulus of elasticity of the lead. If you over crimp, you'll quickly exceed the modulus of elasticity of the lead and the brass will "sping back" some amount when the clamping pressure is removed.
In this case, excessive crimp results in a net crimping force of zero when compared to an uncrimped round. The top of the case was not in contact with the bullet before crimping, and will not be in contact with the bullet after crimping.
My point is, if the case is not sufficiently holding the bullet before you taper crimp, over-crimping the round will not improve the situation. If you are able to apply so much over crimp that the case somehow grabs the bullet enough to hold it (usually by deforming the jacket) you'll end up with cartridges that are inaccurate, inconsistent, and maybe even unsafe.