In my lab I have created a 'Hub & Spoke' network. I want my spokes to reside on the same network address so I can choose between using the physical interface at the hub or a multipoint interface at the hub.
Why would I prefer one technique to the other ? As I see it 'one' physical interface is taken up by both methods !
Also with both techniques I can achieve solutions using inverse arp or static mappings.
What am I missing here ?
It could be a migration to point-to-point or just a design requirement.
For instance, per your requirement, you want some spokes to have the same IP subnet as the hub, well - then you configure a multipoint subinterface.
Later on, you want some spokes to have a different subnet, then you configure a point-to-point subinterface along with the multipoint.
That's the flexibility from having that option as opposed of having point-to-point as the only subinterface type available in frame-relay.