In order to create more than 4 partitions using the DOS/MBR partition layout, one typically uses extended partitions. Basically, this means that the last primary partition contains another partition table in its very first block (which has the same layout as the true MBR).
However, instead of using all 4 partition table entries, to define additional partitions, only the first two entries are used (EBR on Wikipedia).
This means that defining many secondary partitions would result in a linked list of partition tables instead of a tree structure.
My question is, why was this approach chosen as I don't see any advantages in this design.