I like the look of it. But it will only work well if the degree of movement is small for reasons PhilF mentioned. How small is probably important. Clearance in the hole is probably important, too.
Breakout (if any) would be at the bottom corner of the hole. All depends on the magnitude of the vector force when locked and the slide material properties (CI, I assume). Might be fine, might not.
Also, depending on the hole clearance to the ball, it isn't just movement to the left wall of the hole by the upper ball that might be a problem as the lower ball moves, it is also movement fore and aft (in this profile view) if clearance allows it. These are balls so they slide in two degrees of freedom with pressure from above. Any movement that will allow the top ball to seat lower will occur, so that may not be just movement to the left. Depends on clearance, movement, etc.
But mental experiments take second place to real ones. I'm a great fan of reality! It wouldn't be too hard to mock this up in scrap to intended dimensions and see what happens with this kind of lock.
