Flow and CFD happen to be one of things I do for a living, that's why I'm quite interested in how it works, although my skill level is borderline retarded.
Given the Reynolds number of typical exhaust gas (composition tikamed with AGA-8) and the bend factor, I'm not sure how it would increase velocity via the swirl. I guess the velocity would increase from the restriction but the swirl would create more turbulence, sapping it of kinetic energy. But I guess I'm not qualified enough to give a final answer, given the shape of the thing.
Just dreaming of a good exhaust after my friends at CEEFC threw around the idea of a few of us quitting our jobs and building high performance Inconel manifolds and titanium exhausts.
The best way to create backpressure for a system without turbulence or disturbing the pulse or lamellar flow is with a venturi type of restriction. Preferably cast or machined rather than rough welded.
velocity will increase for swirl bullet(for low rpm), same like d idea of swapping straight flow to sflow muffler