Sound designVocalsTalkbox Versus Vocoder
Both make an instrument talk. One uses your mouth as the filter, the other uses your voice as a control signal.

A reese is not a preset, it is an interference pattern. Two detuned saws moving in and out of phase, and a sub that has to stay out of it.
A reese is what you get when two saw waves are detuned against each other and left to drift. The movement everyone recognises is not modulation and not an effect. It is the two waves falling in and out of phase, cancelling and reinforcing on a cycle slow enough to hear. Once that is clear, every decision in building one gets easier, and so does every decision in mixing one.
Start with a pair. Every extra voice adds another set of cancellations, and past a few of them the cancellations average out into something wide and static rather than something that moves. If the sound is not moving with two voices, more voices will not rescue it.
The detune amount sets how fast the sound churns, because the rate of the movement is the difference between the two frequencies. A small detune gives a slow, heavy drift and a larger one gives a faster growl. There is no correct figure, and the one you want shifts with the note being played, which is the first thing that makes a reese hard to keep consistent across a bassline.
A raw pair of saws is mostly top end. A low-pass filter brings it into bass territory, and slow movement on the cutoff is what turns a static tone into a part. Much of the character people credit to the detune is actually the filter moving underneath it.
Distortion adds harmonics, and the harmonics of a moving signal move with it. After the filter it thickens what you have already shaped. Before the filter it fills the range you were about to remove, and the filter then spends its time undoing the distortion you just added.
The interference that makes a reese interesting is also what makes it unreliable at the bottom. Two detuned waves cancel periodically, and a cancellation in the sub range is a moment where the low end is simply not there. On a large system that reads as the bass dropping out on a cycle, which is heard as a fault rather than as movement.
The fix is structural rather than corrective. Keep the reese above a crossover and put a clean sine underneath it for the fundamental, mono and undetuned. The reese carries the character and the sine carries the weight, and making that sine audible on a small speaker is a separate job with its own answer.
Most reese patches are wide, and the width usually comes from the same detuning that creates the movement. Summed to mono those voices cancel far more completely than they do in stereo, so the part can lose a large share of its level on a club system or a phone.
This is worth checking early rather than at the end, because the sound you designed and the sound that survives summing can be genuinely different instruments. What summing does to a wide part is the same mechanism at work, and a reese is the most exposed case of it in a track.

PLUNGE covers sub, 808 and reese synthesis in one engine, built around low end that stays intact.
Guides
You may also like
Sound designVocalsBoth make an instrument talk. One uses your mouth as the filter, the other uses your voice as a control signal.
Bass designMixingPhones and laptops have no sub driver. If your 808 only exists below 60 Hz, most of your audience hears nothing at all.