Looks like pretty normal behaviour, except for the missing rpm signal. A missing rpm signal ought to result in 100% fan speed as well; this is a designed safety feature. The same could happen due to a disconnected pwm wire; signal=off is a much safer choice than a signal=on type. With both a 'malfunctioning' system would give you a bit of unnecessary noise, but never the risk of thermal issues.
Guess you could swap the brown wire to the yellow pin and leave the yellow wire disconnected. This is perfectly safe since the schematics indicate pin #3 (brown) sends the pwm signal to the fan. This would also result in 100% fan speed, but might give an rpm report back. If so then the brown an yellow wires are swapped on this fan and switching them on the connector would give back control plus the rpm signal.
Don't think that'll happen though. More likely the fan's defective or its rpm range is outside that of the one expected by the EC. Did you try with HWiNFO and SpeedFan?