What does "IPC: Frame encode took xxx ms so far. Dropping down the quality to index x." in the log files indicate?

Answer:

In this case, the imaging encoder has determined that too many bits were being sent in the upper portion of the frame (i.e. flooding) which causes the frame to take longer to encode, and therefore appears unresponsive to the user (> 750 ms to encode). During these cases, the quality settings used to encode the lower portion of the frame is then decreased in efforts to get the frame out faster.

This is not a fatal error but just signifying that a user was doing a large update in a low bandwidth situation. (Low bandwidth can occur during times of network congestion and can be a reason why PCoIP sessions can be bursty). When most of the screen has changed, it needs to do a large update. The imaging encoder tries to get the large update data out using the available network bandwidth. If the data being sent is low due to a low bandwidth situation, the result can appear as screen lag.