I think proper solutions are more complex than breath controlled = volume. For a start the commonplace nature of keyboards and pads means that initial velocity of something being hit is the norm for controlling the volume of the note/sample to be played, and initial value is not much use for a breath controller where the value is going to change over the course of the note. So normally with software instruments designed to be controlled by breath, the breath signal is set to have an affect on certain envelopes or various other parameters of the instrument, well beyond volume. If the volume is allowed to change over time via breath, rather than just setting the initial note velocity, then it may be ok, but it may still not be enough to give satisfactory results. What you really need is to hook it up to a sampler that does clever stuff based on breath pipe data, eg add aspects of other samples to the sound to simulate how real wind instruments sound when breath changes. In the case of synths, you ideally have a synth based on the physical realities of how the real instrument makes a noise, and tie the breath controller to the appropriate part of the synth. There is a real nice reaktor patch I posted about called Silverwood which can be made to react to the pico breath controller and the results are way different to what I would expect from all but dedicated wind instrument software. Other synths can sometimes be wonderfully expressive using breath if you experiment and find a suitable synth parameter or two to tie to the breath controller, but many synth presets dont translate to breath in an easy or rewarding way, and I imagine the problem will be even worse for sample-based instruments.
Is there a good reason why the pico uses a midi cc other than the standard one for breath? Granted there arent so many software breath instruments around to cause too many problems, but I had to hack that reaktor patch to listen to the picos breath because it wasnt using the standard midi cc for breath.