One other point about having a quantizer: you don't have to use it with a sequencer.
If your quantizer is capable of loading different scalar patterns, you can restrict it to only the specific pitches you want to appear psuedorandomly, and then you can feed it things like complex voltage curves, sample-and-hold CVs, LFOs, etc, and you'll get randomish behavior each time the quantizer is stepped by the clock or gate/trig sequencer. Plus, a few quantizers take up less space than most full-on CV sequencers, so if the goal is to create these sorts of stochastic patterns, you can use a few of them to create shifting polyphonic harmonic structures off of various psuedorandom inputs. Think of this as being sort of like using a sample and hold, but with discrete pitch results as opposed to the randomness offered by noise signals.
One other useful point: you can also use quantizers as analog shift register stages. For example, if you wanted a four-note arpeggiation, you would patch quant #1's CV out to #2's CV in, then #2's CV out to #3's input. Quantizer #1, of course, would be your controller CV. So by multing that controller CV out separately, plus all of the quantizer CV outs separately, then "clocking" the quantizers with a multed trigger from your controller, you'd then have four CVs with canonically-related pitch outputs, and the "canon" would step each time a note on the controller was played.