"The rotary bolt opens due to pressure on the slopping surface camming it around, sort of like squirting a watermelon seed."
The slope in locking surfaces is around 6 to 7 degrees. I find it hard to believe that the forces tending to throw a double gun open are so great that they can force such shallow angled surfaces to act as cams.
If those forces were so great doubles would not be able to fire when held by two layers of scotch tape (per Gough Thomas), or no locking bolt in place (Greener).
More likely is the design that locates mass in the "wrong" place so that the inertia during recoil gets the parts moving. Note how Hoof reports fixing his opening problem by replacing weakened springs. This echoes my experience with this problem.
Slow motion videos on Youtube show that the opening levers and hammers (of hammerguns) oscillate under recoil. And let us not forget that inertia in Benelli autos is strong enough to compress a stout recoil springs that cannot be hand compressed.
The inertia of a bolt acting axially with the bores is subject to the forces you mention. The rotary bolt has no motion which would be moved in an opening direction by inertia. I am not an engineer, but seem to recall that around 7° is the point where friction is about equal to applied force. I believe one problem with rotary bolts is that the angle actually increases across the locking area due to the decreasing radius to build in the wear compensating taper. If the bolt is not absolutely perfectly fitted it can end up with more pressure out to the right side of the lug where the taper is greater. Elmer keith once reported on an Ithaca 10ga NID which he said was the "Most Automatic Gun" he had ever fired. Upon pulling the trigger it fired, opened & ejected the empty. He sent it back & they refitted the bolt & cured it. They "Did Not Change the Moment of Inertia" by simply refitting the contact area of the bolt. A stouter spring of course increases the forces necessary for opening the bolt whether it be from inertia or otherwise.
