No, there is no running code for the API, and I understand that a piece of running demo is needed. On the other hand, when I consider this as a target for my activity, I have problems to identify an attainable and significant goal: an API that allows to configure a ping between two OpenStack VMs is meaningless, but expensive. Usability is a key factor that is not demonstrated by one use case. Performance shoudn't be an issue: it is an interface to define the infrastructure, not the infrastructure itself.
I read the presentation by Parpaillon, and the idea to represent in XML the monitoring extension came to my mind: the question is if this is considered as "running". Or, better, to what extent this operation might be considered at least as a "proof of concept". Who is going to be convinced by that?
This is probably the next question: given that the problem is defined, and that the approach to the solution is convincing (110 views during one week makes one result), we (as a group) should define what kind of result would be able to foster adoption of a solution, which is the final goal. And I need help on that way.