Would stress really go up? Sure you have a large lever arm when the piston rolls over the nose of the cam and starts pushing it down, but it seems that there is a large bearing area available. In a standard IC engine all that force is transmitted through a half inch wrist pin and a two inch rod journal.
I just spoke with another engineer here and apparently Revetec has been hawking this engine for a couple of years without any real bites. Cool tech, high power density, but expensive.
On another note, I'll bet things could get interesting in a hurry if one of these engines develops a misfire.
You've spotted the weak link in this design.
There was a very similar design about thirty years ago that used the same double ended piston idea (six pistons, twelve cylinders) and went this design one better. The cylinders were disposed in a circle around the main shaft like a revolver, in the center instead of two counter rotating cams there was a single swash plate similar to an air conditioning compressor. (a swash plate is a disk which intersects the shaft at an angle and wobbles back and forth as the shaft rotates, the angle of intersection being the stroke) The plate used in this engine was trick in that it was designed to provide two strokes per revolution so it was a bit more complex then a flat plate on an angle. The trick allowed the output shaft to run at half speed compared to the pistons. This allowed a simple face cam track at either end to directly power the intake and exhaust valves from the main shaft without timing gears (think double overhead cams WITHOUT the cam shaft!). Ignition could be driven from the same cam (at the time breaker-less ignition was just make a start) No distributer was required to run a standard four cycle engine with twelve power strokes per revolution.
All told you had a twelve cylinder engine, shaped like a thirty gallon drum with about 1/4 the parts count of a V8 with about the same displacement and a torque curve that started down near your socks. The built in two to one reduction made it a natural direct drive of aircraft propellers and as I recall one of the light plane companies was looking at it. (Cessna maybe, I don't remember for sure) I've never seen it since.
Back to the RevTec: Actually, the two cams seam to me to be unnecessary. The mechanism is desmodromic, that is the distance measured across the cam is a constant as it rotates (conjugate surfaces). The result is that piston motion in either direction will cause the cam to rotate. Key point: there is no method to adjust for "lash" or clearance except to dial it into the dimensions. ANY wear on the cams will cause the worst case of "rod knock" you've ever heard. As Nathan has pointed out, this is a highly loaded area in this design and therefore adjustment mechanisms would have to be robust in the extreme yet very stable over the operating temperature range. A formidable task! That is very probably why that "revolver" engine never saw production as it shares the same weakness.
Regards,
GtG