Could the rotors be modified to perform better at auto rotating and maybe less at powered performance?
In general, regardless of the number of blades, the more blade inertia, relative to blade area, the better for autorotation safety because the inertia buys you more time to establish the autorotative glide before blade stall when power is removed, and more time to settle into a landing at the end of the glide. The downside is more sluggish response to power application and maneuvering inputs on top of the added weight itself, so there are the usual tradeoffs.
The Robinson R22 started out with very low blade inertia for maneuverability (Robinson did not initially intend the R22 to be a trainer), but blade weight was increased later in production to make the aircraft more forgiving, to help with the blade-stall-following-power-loss accident rate. In the earlier R22's you had under 2 seconds to lower collective pitch following a power loss before rotor RPM decayed enough to get blade stall, and if you weren't fast enough, that was it - you were trapped in a falling object (that's not as bad as it seems - if a car stops or swerves in front of you, you have a hell of a lot less than 2 seconds to react to avoid being killed, don't you).
In large helicopters or smaller machines with high inertia rotors like 2 blade Bell machines, you have oodles of time to establish the glide and settle after the flare. A Huey can be lifted into a hover, held for a couple seconds, and settled on the ground just on rotor inertia alone.
You see a somewhat similar issue between regular airplanes and flying lawn chair ultralights. When an engine quits on an ultralight, because the thing is big with lower mass, it slows down pretty fast and you have to aggressively shove the nose down right away to maintain flying speed compared to something like a Cessna 150. And a helicopter rotor is just an airplane going in a circle.