Food for thought:
If you enter a climb by flaring in autorotation then the aircraft will have some amount of inertia carrying it away from the force of gravity. While in the "flare" effect there is the required airflow through the disc from beneath producing an autorotative force and driving the rotor. Should you then apply forward cyclic to maintain some acceptable airspeed (which you need to complete the autorotation) you have to wait until the aircraft stops climbing from this inertia and commences a descent before you can re-establish autorotation. In addition, the application of forward cyclic will unload the rotor disc and your RRPM will reduce from coriolis effect. The net result being a potentially significant loss of RRPM, close to the ground, which in my opinion defeats any logic of doing the manouvre in the first place.