The thing that want to make clear to myself is:
Do WISMOs (or SDKs for them) allow using internal speech codec in user applications?
In my opinion, that would be an attractive feature to allow user application to compress voice data, coming from MIC, input with hardware (or OS APIs) and transmit it over Internet. And reversed. Some kind of IP Telephony. Even if hardware configuration does not allow to use MIC and SPEAKERS, there still remain ADC and DAC. But then a problem arises:
declared ADC perfomance is higher then 60KSps (even 90KSps in 2686) and all that can be used is 10s update period. Maybe there some special APIs for ADC and DAC?