Using a second Kinect is primarily intended for improved accuracy, not range. For that purpose, the capture areas of individual Kinects should overlap for an actor to be visible simultaneously from two viewpoints. Another reason for overlapping is that otherwise they cannot be calibrated using the current method (with slim rectangular object).
For the same reason we do not support more than 2 Kinects. You can find some discussion on this subject here:
http://forum.ipisoft.com/viewtopic.php?f=2&t=6245