API 2 (Support for pitch and improved parameterization)

The support for pitch was added to the SAPI backend (currently the only backend) which required a rework of the SAPI usage to construct XML fragments for the ::ISpVoice::Speak calls. This also incidentally simplified planning of pitch, speed, and volume for speech fragments by way of decoupling the setting of parameters from immediate calls. This will be the final commit for speechio API 2, as it will be superceded by speechio API 3, which reworks speaking jobs to enable easy prerendering and which uses its own audio output to improve responsiveness to an extend acceptable in realtime interfaces (such as video game interfaces). <3