How would you design a distributed decision-making system?

I’m toying with an idea of a DMS (as per Q). Allowing all members of a group to have visibility to both a decision and its factors or data. Any decision can be amended given secondary feedback. But often there are situations where a decision needs to be made in that moment. So, hold the initial decision as current until further modifications can improve its efficiency. The individuals involved and affected by it get a say as well—replacing static law by dynamic distributed mediation.

Visibility arises from sharing what the individual sees to the best of current technological advances. Video/sound, documents in real-time (safeguards forgeries). As in, the rest of the network can witness directly (increasing the number of witnesses) and verify recordings of an event experienced by a member of the group in the environment,

This discussion is not about whether or not such systems agree with you, but how to best design them. With differing levels of application. (As in current tech, low cost up to future tech, high cost).

A couple of examples in a similar direction: http://earthengine.googlelabs.com/ and the haiti tracking websites for stress calls (urls?).