Why would that make a difference? Surely the splines take the load, regardless of locknut torque?
The larger sprockets are held together by bolts / rivets, so the load applied to one is transferred to them all and reacted by their collective contact area with the splines on the freehub.
It seems reasonable to assume that if the locknut is tightened down sufficiently, load applied to the smallest sprocket would be reacted (at least partially) by its interface with the rest of the cassette - this load being spread across their collective contact area with the splines and lessening the amount of load carried by the interface between the single small sprocket and the splines.
We can see the benefit of sharing the load across multiple sprockets by the marks on the hub pictured - despite the fact the larger sprockets exert more torque on the hub for a given chain tension, they've still caused less damage than the one smaller sprocket.
Pretty disgusted that this is considered normal - IMO this is not an acceptable outcome and either the design or material needs to change to prevent this. Of course steel would be fine, were it not for the previous weight-saving demanded by the marketing department..