A couple of pages back, one poster pointed that there isn't anything wrong with the Agere chipset its the people coding for it. To summerise his point, if the Agere chipset can handle real time video without any issues then real time audio should be easier to handle since there is less information to push through the chipset. If that was the case, then the 3rd party is doing something wrong, not the chipset.
His analysis sounds more logical than it actually is - at the end of the day, they've switched from a chip that works 'perfectly' with 'everything', to one that (we/they know) doesn't.
There are loads of different chipset manufacturers making components that do the same thing as one another - they certainly don't all perform the same.