Large Crowd backups are slow to import/export. It would be good to go over the import/export code to ensure that we are reading and writing the backups in the most efficient manner.

In particular, we should update our import process to adhere to the dom4j guidlines: http://www.dom4j.org/faq.html#large-doc