There is always a gap (there has to be so the gun can be opened) and every time the gun is fired the latch tries to pull apart and batters itself.
All previous revolver latch designs to the best of my knowledge were non-tensioned. A yoke simply wrapped around a post, or two. A spring kept the yoke in place, but the post (frame) still had some play. These designs battered/wore fairly quickly.
Break-open actions using tensioned latches do not have this problem when designed properly. An African game cannon has enormous bolt thurst, yet a good one will last many shots. This is because tension in the mating surfaces prevents a gap from opening that can lead to wear or battering. A latch/lock does not
have to have a gap to open/close, just a spring.
Improving upon the old trunk-latch style frames would not be difficult, but it would be expensive. Unlike a modern DA/SA, a designer doesn't have 100+ years of production tricks to keep costs low and the product consistent. But the concept is far from impossible. The trick is designing the locking surfaces to remain compressed together harder than the forces shaking them apart.
Heck, my half-baked auto-revolver design can theoretically handle 90kpsi proof loads because I centered the latch pivot over the barrel (like every break open rifle/shotgun). The loads holding the action closed are so low that the cylinder axle can be used as the latch member, like an old Colt SA. Maybe I'll finally be able to get a proof-of-concept-model made by hooking up with
these guys--or just panhandling over at Kickstart.com
TCB