Have most backends use their own default speech rate, volume, and pitch instead of just 50%

[emassey0135]

The synthesizer metadata should also specify these defaults, in case applications want to have these defaults in their configuration UI. This would let the OS or Speech Dispatcher settings be the defaults instead of it being the same for every user.