Try thinking about it the other way: it’s not a pitch bend, or an ‘octave’ shift: it sets the actual pitch of the oscillator. Ignore the keyboard for a second. A VCO is just emitting sound constantly, oscillating at a frequency. The “pitch” knob on an oscillator like is on the 0-Coast sets the pitch its at.
We can alter the pitch without twisting it by adding control voltage to the Volt per Octave input. We put a voltage in, which represents a certain amount of pitch: one volt represents one octave above the base pitch; two volts two octaves; and so on - a few millivolts represents a tiny increase in pitch. And yes, negative voltage would shift it lower.
So we tune the base pitch, and then feed in voltage to make tunes.
Now - when you plug a midi keyboard into the 0-coast, you’re effectively converting those midi notes to CV internally, and that CV gets added to the pitch that’s set on the pitch knobs.
@Starthief’s point that the pitch knob, the CV, and midi are all methods of transposition is a really good one: essentially, all the voltage gets added, and fed into a thing inside the oscillator called an ‘exponential converter’, and ta-da, the pitch of the oscillator is set. (This is why you can add voltage to a CV input on a VCO to transpose a sequence, for instance).
The 0-coast is an interesting beast in that it’s less rooted in straight up keyboard-based instruments than the Moog - that’s what’s known as the ‘west coast’ influence coming in on it.
Sorry if I’m going on about this, but you’ve stumbled onto a fairly core concept in the land of voltage-controlled synthesis: the relation of voltage to oscillator control. The 0-coast is good at laying this stuff bare, I think. I certainly wouldn’t want you to ‘not ever touch the pitch knob again’ - it’s a control on a synth, and the designer has placed it prominently on the panel, that means it’s got to be for something, right?