...my current issues is this: I've taken a cello voice, tweaked the reverb and the release and made it perfect for what I need. I've stored it (and even turned off and on the machine to make sure it's stored). However, when I go to use the same voice (user 1 A06) it's back to the dry factory setting. Anybody know what gives?

There are actually two levels of editing on the MOXF synth... The surface level (or quick access level) and the in depth (Voice Edit level).

Why these two levels exist is to serve the two primary uses for the synth: Live performing and Multi-track recording (music production).
For the Live performing use case you have the QUICK ACCESS parameters found on the Knobs on the left front panel. These apply "offsets" to the data that is stored within the Voice itself. Among the Quick Access parameters are "Reverb Send" and AEL "Release" (the two you mentioned you edited). Those are applied as OFFSETS to what was originally stored in the VOICE. The Release parameter for example, adds or subtracts the value depending on if you turn the Knob clockwise or counterclockwise. This offset is applied to all 8 Amplitude Envelope Generators that make up a MOXF Voice. Remember, a Voice can have as many as 8 Elements - each with its own oscillator, filter and amplifier system. Imagine how hard it would be to simply make the sound's release longer - "quick access" means you can turn the one [RELEASE] Knob and it is the same as going into deep editing and having changed each of the 8 AEG's manually.

So things like Cutoff, Resonance, Attack, Release, Pan, etc., are "quick access" (knob'd) parameters that you can get to quick on-stage. They are usually not stored, but the MOXF allows you to tag these to the VOICE (as extra stuff).

However, these are stripped OFF when you then select the VOICE in a PART (either a PERFORMANCE or SONG/PATTERN MIXING). Only the actual (original) stored condition is recalled - the extra stuff is stripped off.

Summary: Offsets are applied to the VOICE at VOICE mode level and can be stored in the VOICE in VOICE mode, but they get "zeroed" out, when you select the VOICE for a PART. They are extra parameters tagged to the VOICE but not deeply edited into the Voice. Like the Arpeggio Types, these are temporary baggage. They are not really apart of the VOICE, they are extra parameters tagged on to the VOICE... Whether or not they "travel" with the VOICE when you use it elsewhere is the big question!

That is why, when you place the VOICE in a SONG MIXING setup, the RELEASE (offset) you made is stripped off. It's not really deeply edited into the VOICE, just an offset tagged onto the Stored Voice settings. Normally, you would have to go into deep editing and change the AEG RELEASE Value of each of the VOICE's individual ELEMENTS... and store that into the VOICE.

That all said, if you would like to include these OFFSETs you made back in VOICE mode, you can have the MOXF *copy* those to the MIXING PART for you!!! _ Yes, the MOXF can go and get these parameters, if you would like.

Here's how:
Turn the "Parameter with Voice" setting to ON prior to making the assignment to the PART.
"Parameter with Voice" is a function that copies the Offset parameters with the Voice when you select it in a PART. So any offset values applied to the VOICE, including Arpeggio phrases, Filter and Amplitude functions, etc., will be brought along when the Voice is selected for the PART.

Press [MIXING]
Press [EDIT]
Press PART [1]-[16] to view a PART
Press [F1] VOICE
Press [SF1] VOICE
Here you can set "P.With Vce" = ON
Then select or reselect the VOICE to 'copy' the offset parameters to the PART.

Please recognize this does not include the REVERB SEND amount. Reverb is now a communal effect that all PARTS in the MIXING will share. An individual VOICE cannot bring along its own Reverb - it is impractical - beside the ensemble is assembled in a single room. So reverb send amount will be very much affected by the other instruments now sharing the selected Reverb chamber (common) for the entire MIXING setup.

Hope that helps.