The ball valve is a nice idea. It would be larger and heavier than an iris, but you're right in that it wouldn't wear much and would be very simple to manufacture and operate.
Here's an rough sketch of an idea I had for a dual valve system that might not need mechanical actuators. If done right, I think the changing pressure inside the combustion chamber should be able to move this valve(s) by itself. It wouldn't contain the exhaust before letting it out, but it also doesn't have any extra complexities. Basically a half way point design, I guess. You get some of the advantages without some of the disadvantages.
The inlet portion of the valve has a larger area and so should be affected more by the pressure than the exhaust portion. The increase in pressure do to combustion should shut the inlet, thereby opening the exhaust. Decreased pressure as the exhaust leaves should open the inlet, closing the exhaust. Ram effect should help open the valve in the intake stage.
The question is, will a sufficiently strong vacuum be able to form before the valves switch to intake position...