I was thinking about previous beta tests the other day. One of the most successful beta tests I participated in was for Europa Universalis. Part of this success was due to one of the beta testers coming up with an excellent way to assess the AI's ability. Fortunately for EU, this ability was already written into the code.
Simply put, EU provided a way to disable fog of war, explore everything and watch ALL world events as they happened via a message system. The beta testers turned on the game, picked a country no one usually bothered, and accelerated time to maximum. End result was that we could see how well the AI was doing against itself. We could see when the first colony was built and where. Who started the first war and how successful that war was. We could view each empire's treasury and current technology levels. The list goes on and on.
I would like to see this ability installed for the next phase of beta. That way we can help you better assess the AI's ability to deal with us and itself. Basically, it would entail cheatcodes to clear away all unexplored areas and fog of war. Another code to be able to see what each empire is doing at any given time (i.e. select their planets and see what the colony is like, find out what they're researching, etc). And lastly, a text message system letting us know when certain types of events happen. Enemy ships built, enemy planets colonized, enemy research completed... basically everything we already know about our own empire via the drop buttons. These codes would have to be in the game by beta 3 though, otherwise we wouldn't be able to use them during the phase we really need them.