I don't think it can, Dennis. I have got all the manuals, and have read them (I know... but someone has to do it!) and what we are talking about here is more like the old 'unison' mode from the old Jupiters, and the like.
If you hold down ONE note, all the oscillators sound (fat!). If you hold down two notes, half the oscillators go to the high note, and half to the low note. Hold down three, it gets split to three, four to four, etc..
About the only way this seems even slightly possible is to have each voice as a monophonic sound, and have High note and Low note Priority to them. But this only seems to work for TWO sounds. Play anything more than two notes and no further split can happen. Not to mention, as a monophonic sound, should you lift one finger and leave the other sounding, the note 'lifted' will jump to the note still held.
It's a far more complicated thing than it sounds, Dennis. It really IS more to do with the voice allocation routines than it is to the voice programming.
So Diki are you are saying the PA is like the old jupiters unison mode???? if so, you are way off the mark!!! and again if so, I don't think you really understand the power of the sound programming on the Korgs!!
For a start, voice allocation priority (which is what it seems you are talking about in this second post) is on ALL synthsizers, no matter make model or cost. Even the G and this add on card you are talking about...there are just so many DCO's that can be physically built into the PCB's on these things...
The Korg has 16 independently programmable oscillators for each single sound patch...as I said,
each oscillator can have it's own dedicated adsr parameters, triggering parameters, velocity parameters and so on..which is rather more than just the old "unison mode" doncha think?

D