I think the most satisfactory result will be to use ECN-2 film as an interneg material and print onto ECP-2. I intend to do this someday, but the minimum orders are rather expensive, and at the moment I'm a bit tied up doing emulsion making.
As for B&W (and with any duping), contrast control is the main challenge. If you start with a normal pictorial negative, it will have a contrast of maybe ~0.7. If you contact print it (or take a photo of it) onto the same negative material developed the same way, the resulting positive will have a contrast of ~0.49. The contrast curves multiply:
γtransparency = γnegative x γprint material
γtransparency = 0.7 x 0.7
γtransparency = 0.49
This will be unacceptably low contrast. The target aim for a transparency should be a contrast of about 1.7-1.8 ish. So, if you've already got a negative in hand, you'll need to print it onto a material with a contrast of about 2.5:
γtransparency = γnegative x γprint material
γprint material = γtransparency / γnegative
γprint material = 1.7 / 0.7
γprint material = 2.5
A specialty film like ortho-litho should work with the right developer. Or maybe a microfilm material? It will probably take some trial and error. Sounds like a cool project! If you do go down this route, please post your results!