(1) UK Gyroplane fatal accident rate is 1 per 6,000 hrs, compared to, about 1 per 50,000 hrs for light helicopters, microlights and gliders. By any benchmark except financial, Whirly is far better off in her Robbo - although it's not QUITE as bad as wibbly suggests.
(2) Let the rotor speed get too low on any teetering rotor system (Gyro or Robbo) and it is likely to end in a fatal accident. It simply happens that the handling errors leading to this are subtly different in each case.
(3) Homebuilts are subject to rigorous inspection during build, and like any other aircraft during use. Nothing unusual in that. In both, it's the breakdown of the inspection regime that tends to cause any mechanical related accidents.
(4) Never seen a 447 on a gyro, too low powered at 38hp. Mostly 618s, 582s, 462s, etc - 60hp+ engines.
(5) No doubt that Ken Wallis started with a modified Benson, but he came a long way from there. Equally no doubt that both post-dated Juan de la Cierva by some considerable time. Cierva deserves much of the original credit, although subsequent work by Glauert, covered most of the theoretical aspect, and practical designer like Benson and Wallis brought things forward a lot post war. The big change Benson made was to stick the prop at the back.
(6) Flare dammit, you obviously know a fair bit about the subject, but winding up whirly is hardly the way to establish your credibility. Whirly, being wound up by an obvious wind-up-merchant like FD doesn't help yours much either.
(7) There's a gent, a reasonably regular poster on Pprune, who has done a great deal of theoretical work on gyroplane safety. Perhaps he'd care to chip in?
(8) A personal gripe here, I think the problems with the safety of the Benson aircraft was down to a lack of theoretical investigations, particularly taking into account Glauert's non-dimensionalisation of gyroplane stability in the 1930s and 1940s, in the basic design. Adding in a couple of cables to reduce flexibility is not the solution. The gent I mentioned in (7) has gone a fair way towards solving this, although I've yet to understand why he doesn't non-dimensionalise his own published theories.
G