Why does 8086 processor adopt the segmentation mechanism?

Anyone who has learned the Assembly knows that the 8086 processor uses a segmented management mechanism for memory. Why is this design necessary? How nice it is to directly access the memory using a linear address.

In fact, Intel engineers will never be so stupid. Of course, there is his theory:

First, the address bus has 20 IP addresses that can address 1 MB of space, while the registers are 16-bit and can only address 64 KB. The extra space cannot be wasted. You know that in 8086 S, memory was very expensive.

Therefore, Intel engineers have come up with a segmentation management method: segment address X4 + offset address.

In addition, segment management is conducive to code organization and effective management of memory. For example, you can isolate code segments and data segments.

During the learning process, if you copy the learning as described in the book, it is neither easy to remember nor integrate the knowledge.

Why does 8086 processor adopt the segmentation mechanism?