When Casull was first developing his .454, he used duplex and triplex loads - but he didn't mix the powders, he loaded them so they'd lay in distinct layers and seated the bullet so as to compress the powder charge, and keep the layers separate. The advent of new powders made this technique unnecessary to achieve the velocities he wanted. I've heard of using a small charge of smokeless right above the primer to facilitate cleaner ignition with a large black powder cartridge, but not being a BP shooting myself, can't say if it's safe or effective.
I wouldn't seriously consider blending different powders, even if they're similar - for example, even mixing IMR4831 and H4831 would be a "no go" for me. You really DON'T know what you're going to get, and some unexpected non-linear effect could cause a real problem. Ditto with layering different smokeless powders in a modern case, either rifle OR pistol.
The only "safe" mixing or blending technique I know of is to blend two different lots of the exact, same canister powder together. Say a person purchased two cans of IMR-4350 to load up a bunch of ammo, and then found out his batch would require 1 1/2 cans - but what he bought is 2 different lots. I don't see a problem with blending them together and working up a new load, especially since canister powders are controlled pretty closely as to burning rate already.
As Jim Watson mentioned, another "duplex" technique didn't initially use different powders, but rather a tube inside the case to direct the primer ignition to the front of the powder charge. Old Elmer Keith - when he was working for the military - was inspired by its use in artillery and tried this in .50 BMG, getting a couple of hundred feet per second more velocity with normal pressures. IIRC, cost and problems if the gun had a slightly oversize chamber prevented further development for small arms.