I'm sorry if I gave the impression that their work wasn't valuable. It was, I'm glad it was done, and I hate that the drive 'disappeared' for so long. I'm glad it's being revived, too. And I agree that designing/building a safe, reliable gearbox is by far the toughest thing to do to make *any* alternative engine work on an a/c.
But to say that a 'soft' system won't work with a rotary simply because *they* didn't get it to work flies in the face of what engineers tell us, and also in the face of empirical evidence. I don't doubt that they had all the problems you describe. But designing around *every conceivable* configuration & condition is pretty difficult, and often, not required. The trick is to know where the issues lie, and be sure that they're avoided in operation.
You mention Dave Leonard torching RWS bearings, but fail to mention the torched bearings in the Powersport time to climb attempt at Sun N Fun a couple of decades ago. I don't consider either incident relevant to the discussion, since both were outside of normal operating parameters and neither involved torsional resonance issues. Throwing out such examples just muddy the waters of legitimate research. (BTW, I don't want to fly a car transmission, either, but the point is, it worked for hundreds of hours without any problems until one day it got operated in a condition far outside normal conditions.)
Consider the certified world. Many (most?) certified aluminum constant speed props on certified a/c engines (hundreds of thousands in operation for decades) have placarded yellow arcs on the tach where the engine cannot be operated continuously, and they're there because of resonance issues. The Sensenich fixed pitch aluminum prop for 160 HP Lycs on RV-x's (thousands in operation) is placarded against operation above 2600 rpm, also due to *high frequency* resonance issues (engine is designed to operate at 2700 rpm continuously for 75% power at altitude). All have been demonstrated to fail if operated incorrectly, but because they're operated in compliance with posted restrictions they're safe.
The Powersport research was a for-profit operation, so for good reason (for them), the full body of research has, to my knowledge, never been published. If it was published, we might see any number of factors that show their earlier designs could be operated safely by simply avoiding continuous operation at resonance points, or avoiding the use of heavy metal props, or changing the compliance of the soft coupler, or adding mass to the flywheel, or...etc etc.
Charlie