Retrieve data for the UDW / Content tree and sub-items list using the filtering API instead of the search API.

The goals are to:

• avoid index sync errors that may occur
• avoid the search backend bottleneck

This will require that a remote (REST ?) API is implemented for that API.