Dear Figure of Merit,
Unlike then Pope, I don't even pretend to be infallable! Sometime it seems that way because I have seen the exact issue, measured it and experienced it, but in this case, I am as much a novice as anyone out there in cyberspace. This thread is another really good one, because from a simple question, real understanding can be pried out. Thanks to Irlandes, who provided the suspect chart, we can do some more examination. In every case, I implore Ppruners to challenge (the ground effect crowd had no such problem!) each of us, so we all learn more.
I certainly agree with you about my fallability, but (with some quibblingness) my original post was a speculation, based on what the limiting factors for rotor hover performance are, and a request for more data to seek the real answer. While the Robinson is probably derated up to the Knee of the WAT curve (the slope change at 5000 to 7000 feet), it is power limited above that. The two slopes of the curve reflect two of the several possible limiting factors:
1) Constant power section - The lower less steep slope from sea level up to about 5000 feet is the fall-off of performance of the rotor due to density altitude, while the power is a constant (this is the area where you are quite correct, the derated engine puts out constant power). The performance loss of the rotor is about 1 lb of lost hover weight per 100 feet of altitude increase (about 50 lbs per 6,000 feet, actually). This is purely the affect of the density being reduced, so the blade must be operated at an increased angle of attack, so the induced power is higher, and some performance is lost.
2) Reduced power section - The slope above the knee is the result of the performance loss due to engine power loss, PLUS the lost rotor efficiency due to density altitude. That slope is about 250 lbs per 5,000 feet, about 5 lbs lost per 100 feet of altitude gain. This is the sum of the 1 lb per 100 feet of the rotor efficiency and one would guess 4 lb per 100 feet due to other effects such as engine power decay and or loss of tail rotor thrust.
If this were not a family holiday in the States I would spend more time with the chart and try to deduce more, but that will have to wait, since a turkey waits for me!