The big factors for a successful autorotaton design are clustered in three places:
1) Good rotor inertia, so the stored energy is high. Dave uses a T/K measurement, which I am sure is the total stored rotor energy (1/2 I omega squared) divided by the total power required to hover (energy/second). The quotient is the number of seconds of stored energy, so it is a measure of how much pitch pull there is at the bottom (although it is not the total number of seconds you have available, since you can't pull the rotor down to zero rpm!). Inertia also helps increase the reaction time available at the initial engine failure. It also helps keep the rotor rpm from wandering during maneuvers while in an autorotative descent.
Pilots love high inertia rotors for auto, and hate them for handling, since high inertia blades make the cyclic sluggish. The principle rotor flapping term is the flap inertia, the brother of the polar mement of inertia we are discussing here. Designers hate inertia, since the loads on the blade/head/mast/transmission feet are due to CF, so these parts get heavier when the blade does. It is safe to say that 1 kg of blade weight might be 3 kg of empty weight, the worst penalty of any part of the helicopter.
Experience has shown that T/K in the 2.5 second range is a very poor structural rotor system, but a nice autorotational aircraft. T/K of 1.0 is poor as an autorotational bird.
2) Low disk Loading helps greatly in two areas, the auto descent rate, and the cyclic flare effectiveness. Low descent rate means less vertical kinetic energy to absorb at the bottom, more time to judge the proper flare height, and more time to find a place to land. Better cyclic flare means more vertical g per longitudinal pitch rate (the rotor has more cyclic "bite") and the pilot can use less pitch angle, and less pitch angle rate to successfully burn off the flrward and vertical speed.
3) Tail wheel or tail bumper to allow landing in the flare, and protection of the tail. This is less important that the above, but far more important than the typical aerodynamic effects like airfoil and tip shapes.
All of the above are compormised down a bunch for twin Catagory A helos, where the probability of a total power failure is very small. They should be more important for a single piston aircraft, and a trainer, where the need to do a touchdown auto is much more normal.