The Hammond demo is polyphonic, each tone has 9 OscillatorNodes for the 9 tone harmonics and 9 GainNodes for the drawbar volumes. I used an additional GainNode for a bit of automatic gain control to prevent too much difference in volume between the various drawbar settings. Each tone has its own ASR envelope (no decay).
In real live a Leslie unit is a loudspeaker box with 2 rotating speakers, one for the frequencies below 800Hz, and one, a horn unit, for the frequencies above 800 Hz. The Leslie has two modes; Chorale, where both speakers rotate slowly, below 1 rps, and Tremolo, where the rotation speed is faster, about 7 rps. The speed of the bass speaker is slightly lower then the speed of the horn. The box scatters the sound around so to say. The simulation has 4 GainNodes, 4 BiquadFilterNodes, 4 DelayNodes, 2 StereoPannerNodes and 2 OscillatorNodes. First the signal is split in its low and high frequency components, these are fed to delaylines with modulated playbackrate for the doppler effect. The output of each delayline goes to a BiquadFilterNode with its detune parameter modulated by the same OscillatorNode trough a GainNode for the correct amount of frequency shift. The output of each filter goes to a StereoPannerNode that is modulated on its pan parameter with the same oscillator, but trough a delay that causes a 90 degrees phase shift, which brings the frequency peak right in the middle between your speakers or headphone.
This unit, also known as scanner, in the Hammond organ is a wonder of ingenuity. It reminded me of the big tuning capacitors you find in vintage radio's. A series of stators are connected to tappoints of the LC sections of a delay line, with the rotor picking up the phase shifted signals at each tap. This website gives a good insight view of the scanner. When vibrato mode is set, all the sound is passed trough the scanner, with chorus the scanner sound is mixed with the unmodulated sound. For the simulation I used a single delay modulated by a 7Hz oscillator with a triangular waveform. The V-modes use the 'wet' signal, the C-modes a 50% mix of wet and dry.