They're not. With the bolt on version the aluminum spacer is inserted into the load path and thus assumes the full load (dead and live) imparted on the vehicle by the wheel, as opposed to the DRS version where the aluminum spacer is acting only as a spacer and the load path remains wheel->studs->hub. If the materials were uniform throughout the assembly then the bolt on kind would *probably* be better (too close to call without calcs), but they're not, and the aluminum spacer is by far the weakest link. Realistically neither is probably going to fail under normal conditions, but in a stress test the DRS spacers are better, assuming a properly assembly and a stud that is up to the task. Hence the ARP upgrade...
One thing the bolt-on version has over the bolt-through kind is a lower sensitivity to faulty assembly, due to the reduced length of the studs that are NOT pressed into the spacer (less potential for torsional flexibility at the spacer).
Sent from my BlackBerry 9800