But why would you need to have Multi Mode on if you only want to use one layer? ... just set multi mode to off, and use it that way
And about the splitting of voices... it's a hardware hardwired setup... when using more than one layer, the voices is ALWAYS split equally between the layers... it's that simple, and not something that can be changed in software (I believe). That is also why you cannot have a split/stack where layer A has 15 voices and layer B has 1 voice... that is how all DSI polysynths with two layers work.
You should basically see the REV2 as two separate synths physically, each having 4 or 8 voices... completely independant, but with the option to use them together as one big synth when not stacked or split (or set to multi mode).