Perhaps it is because for a higher sweep, the distance the air flows over the wing will be greater, ie straight wing chord / cos (sweep angle). Hence more drag?
Then, why is the drag less below M2.1? Why suddenly, above M2.1, the drag gets higher for a higher swept-back wing?
Does this question pertain to the case where the wingspan (distance from wingtip to wingtip) remains the same for all sweep angles (ie, the wing would have to be made longer)?
Or does it decrease as the wing gets swept back?
I assume every other aspect of the wing remains the same. According to the graph the aspect ration is 3, and the t/c ration 5%, for the three wings.
Still puzzeled...