One of the best things the other distributed renders have is that they just need to be working on a same network drive to know all information for that render frame/animation.
I have worked for several years with backburner on a office because they use 3ds Max with vray distributed render and was very very simple to work, we just need to map the network drive and folders on all nodes, set the startup of backburner on each node and turn on all nodes.
Then open the file and start a new job as needed to send the render to all nodes. We just didn´t whait for syncing or have any issues of delay.
After that i found my own studio and because of the cost of maintaining Max licences we have choosen the blender for our render software in 2012.
When we started we start with network render of blender and then we have instaled Loki Render and used until now. Loki just need to now were is the exe file of blender "blender.exe" on each node and use the direct power of each node cpu/gpu just like we left the file saved.
It reduces the time almost be divinding the time per each node so if 1 pc the rener time is 1 hour with to is more or less 35/40 minutes.
My sugestion for crowdrender is to try working just with one network drive were the file is edited and mark on each node that same network drive, if needed inside the addon we can choose were we have the texstures so each time we click on render, crowdrender will scan for each texture used on the file and pick up on that folder to distribute that information for all nodes.
Don´t now if this make sense but it seems to me that crowdrender have been strugling with sync problems.