Its my understanding that the drag induced by two small vortices is less than the drag induced by a single large one.
Winglets have a few functions, sail effect, spanwise flow etc but they also create 2 smaller vortices.
I suspect that the raked design does the same thing?
The sudden increase in sweepback is sufficent to create a small vortice, leaving a smaller component of spanwise flow to effect the tip ie: a small tip vortice.
I suspect some joe-90 has calculated that the extra weight of the winglet, the drag it creates and the decrease in spanwise flow are not as significant as a lighter design with less form drag even taking into account the extra spanwise flow penalty.
Thats my 2 cents worth..... looking fwd to your findings HD