It is available here:
On the left is a screenshot of the face with all the parts on screen.
Most of it is pulled together as intended using CSS, except obviously, the mouths...
NB - this prototype is a button-controlled face that is testing facial expressions controlled by data. It is a test dummy for a FACE API
Although there is no animation showing here, the mouths are intended to be used to creat lip sync to match text to speech, so that Twitter data can manipulate Cloudenstein to talk.
They are traced from pictures of mouths showing the limited number of mouth shapes in use when speaking.