Changes

AudioSink: Enhance description re 'AudioFrame' and 'frameDuration' also using enqueueData() -> 1 AudioFrame

AudioSink: Add getAvgFrameDuration(); API-Doc: Default*: 'frame' -> AudioFrame; init(): Use getAvgFrameDuration() for queue growth and limit.

This change renders buffer dequeueing, growth and limit sticking w/ [ms] values
while getAvgFrameDuration() assists frame count determination.

getAvgFrameDuration() is calculated when buffer is fully filled (queuedBytes / queuedFrames),
i.e. a proper representation to be used to dequeue in duration range
as well as for growth.

This further decouples the frameDuration{->Hint} parameter in init(),
as it is now only used for the initial buffer count (and latency adjustment).

AudioFormat/AudioSink: Use float in seconds for duration to avoid losing precision when dealing with stats, averages etc

AudioSink: Refine context locking where supported, have a more usable universal API interface

AudioSink: Drop getMaxSupportedChannels(), use getNativeFormat(), getPreferredFormat() and isSupported(); Add setChannelLimit() impacting

Add setChannelLimit() impacting getPreferredFormat() and isSupported(),
i.e. to limit channels for e.g. JOAL/OpenAL spatial 3D sound usage.

getNativeFormat() shall be unaffected.

getMaxSupportedChannels() is redudandant -> getPreferredFormat()