Now, the trick IMHO is to identify the game goals (which are derived from the winning conditions) assign the reasonable set of weights to each of them and make sure they are calculated and balanced. But in terms of what to build, when to build, research and generally the direction the empire is going it could be used as a basis IMHO. It didn't work very well on the unit management level either, as tactical game didn't fit very well in to global picture evaluation. That approach doesn't work in Diplomacy implementation as it is very difficult to balance individual weights to provide consistency in the Diplomatic line. Something similar I think is tried in Civ V.
Chess AI is known to look at available options for the next move and assigning weights to each of the static snapshots (positions), which result from that move. Regarding chess AI being pretty decent and pretty much any game AI not much so.
