Collaborative beamforming enables nodes in a wireless network to transmit a common message over long distances in an energy efficient fashion. However, the process of making available the same message to all collaborating nodes introduces delays. The authors recently proposed a MAC-PHY cross-layer scheme that enables collaborative beamforming with significantly reduced collaboration overhead. The method requires knowledge of node locations and internode channel coefficients. In this paper, the performance of that approach is studied analytically in terms of average beampattern and symbol error probability (SEP) under realistic conditions, i.e., when imperfect channel estimates are used and when there are phase errors in the contributions of the collaborating nodes at the receiver.