I had that same idea many years ago, Martin- I wanted to rifle the chokes of a .410 double - though it would be a 'cool' gun for BP loads and a 300gr. slug.

I don't see any reason why not, although I'm not sure what you mean about 'primitive' on the toooling. Use of another barrel (any descent calibre ML barrel should work) as a guide. Cast a lead lapp inside the 'guide' barrel around a mandrel for guiding the rifling head inside the shotgun bore. European barrels usually have a longer "End Choke" than US made barrels,s so the resulting rifled choke should be longer, which is good - maybe 4" to 5" tops - perhaps. thus, the guide barrel could easily be as short as 10" or longer, of course, would be better. Lead lapps on both, one a guide and the other holding the cutter would be the way I'd do it. Lots of measuring to get things lined up for pouring the lapps, of course. use almsot pure lead for the 'guide' and harder lead for the cutting tool's lapp, I'd think. Kind of thinking things out of course, while typing this. Hope the typing's not too bad.

Most older full choke 12's run around .690" in the bore and and .729" to .730" in the main barrel. .040" (40 thou) was the 'normal' full choke, called "40 points of choke".

This will give .020" or so per side of rifling depth to your 'paradox'. The ball used should fit the major bore diameter to within a thou or so, just as a Paradox bullet fits - much larger than the .715" mould you have. When the balls hit the rifled section, they need to engrave enough to take the twist. If you use a cupped wad to hold the ball in the centre, it's possible the wad might transform the twist rate to the ball, but I certainly wouldn't guarantee that. Full length rifled or smooth works fine with cup-wads and RB's, but the cupped wad & ball might merely strip when it hits the short rifled section at velocity.

As for rifling twist, about anything from 56" or slower to perhaps 100" should be fine, I'm sure. Oh course, it would probably have to be faster for slugs in smaller gauges.