I think indoor results may have an effective diversity improvement that's more/different than simple antenna array factor related. With alot of multipath, the arraying of two antennas might improve results.
I have not seen any articles on "improvement indoors with passively arrayed antennas". It would be interesting to see the results though.