Just to answer one of the questions, if you’re making a DDP from this same montage without dither, knowing the DDP process is making it 16 bit (which it always does), you could do the same for the CUE/WAV: force it to 16 bit in the CUE/WAV export file properties (that way it won’t be 32bit). That is, if you’re sure you don’t need dither because the source files are 16 bit, and you’re doing nothing (no gain, fades, edits, plugins) that would require further 16 bit dither. You don’t need further dither if you’re doing NO processing.
Play the files in the montage and look at the Bit Meter to see if only 16 bits are going through to the end of the chain for the entire program, and if so, don’t dither to the CUE/WAV, and set the CUE/WAV render file properties to 16 bit, because that’s how it’s going to the DDP anyway if you’re doing as I think you’ve described. But if you’re doing any processing (gain, fades, edits, plugins), the DDP would need dither too, regardless of the fact that Wavelab’s going to make the DDP 16 bit no matter what.