To terminate.. or not

Yes, surface area makes sense. But do you really get contact over the entire surface? Both the spade and the binding post are probably not as flat as they may seem to the naked eye. The result is contact only on highest spots of the top surface and the lowest spots of the bottom surface. But if you add a lock washer on top of a star washer, now you are guaranteed surface area coverage plus tension force.

You can shield the entire cable to reduce radiated EMI except for the termination. So how do you reduce radiated EMI at this point? The answer may be geometry of the termination. You probably want something flat and wide because it dissipates the magnetic field versus something long and narrow that concentrates the magnetic field. This is one possible reason for a spade over a banana.

Yes, should the washer(s) produce an ideal condition whereby pressure is applied relatively uniformly across the contact surface area - voila! :-)

Vbr,
Sam

Axle - Shielding is a very difficult thing. I assume you're talking about shielding speaker cables from external EMI radiation. I'm not sure that you need to shield speaker cables but in general shielding works a little different than we think. Shielding material of interconnect is non-magnetic but it is shielding cable not only from capacitively coupled but also electromagnetically coupled interference (common mode). EMI in fact IS electromagnetically coupled into the cable but because of skin effect it travels (to ground) on the outside of the cable's shield. Field inside of the shield is zero as long as cable is symmetrical/uniform.
When a cable shield is carrying differential mode current, as in the case of coax, skin effect will cause that differential mode current to flow on the inside of the shield.

Actually, I was talking about EMI radiating out of the binding post and onto adjacent compnoents/cables. Speaker cable doesn't need to be sheilded because it has low impedance and high current.