Often the modulation of the delays A/B is a bit different, either out of phase, or at different depths, etc. Basically, we chose modulations that would create some great stereo depth, but could also create interesting mono effects when summed together. So you may hear this difference on some of the algorithms when comparing one voice to two voices using the same delay and feedback. Of course with the modulation turned off, there shouldn't be much difference, maybe a slight volume bump with A and B together.
