Do you need your polyphonic to be bitimbral?
Imagine split.
Imagine stack.
Imagine shared knobs for the two sounds.
Do you need it to have 5 octaves?
Would you like it to sing all the voices you're playing and just played?
Are you a player? If so, are you okay with divvying up duties amongst dual/triple tiers and/or L, U, or O configs?
Do you want a rich flavor, from the get go?
Was a mouthful of Pomegranates worth the effort?
Do you want greater control of your poly's twists, bends, quivering, adornments, pulsating?
Do you want to surround yourself in dual discrete pairs?
Are multiple recorded passes an acceptable or preferred approach to single take capture- when more than 6 voices or 2 sounds are preferred?
Do you want the knob to always be that function? Always? Period. ... That's very powerful- is my guess.
Would you make use of 4 full-featured oscillators? Do you need it?
Does two different step sequencers on different sounds with room for live playing on top, all on one keyboard, immediately in front of you, excite you?
Or, is splitting that amongst two keyboards a better solution? Each with its own interface instantly available as last adjusted.
Do you want a flavor that is inherently deep, bold? Raw, primitive?
Maybe it's pomegranates and coffee? (And the Prophet 6, dark chocolate.)
Do you need your polyphonic bitimbral?
(Am I asking you, or me?)