So whats new!!.. Its been around for years, infact it was mentioned on the forum over a year ago. Its not even a new idea!
Goes back to the days of WWII, how do you think RDF is done on the roof of police cars, or radio fox hunting done..
There are some benefits of overall system gain by using 2 or 4 vertical antennas, but it is minimal, a single 15dbi omni would out perform 4 x 9dbi omni antennas that were phased...
However, the dramatic improvement is in the "antenna null", suddenly 30db of rejection can be had in unwanted directions..
But prey tell me, now that you have a "super-efficient" antenna system that will steer itself to the direction of the cpe (wanted signal), in a ptmp AP scenario, how do you deal with the effectiveness of time-slots or time-frames or any other form of cpe synchronization when the CPE suddenly becomes 30dB deaf to the AP, whilst the AP is in chatter with a cpe in the opposite direction????