This link https://jazz.net/help-dev/clm/index.jsp?re=1&topic=/com.ibm.team.workitem.doc/topics/t_importing_work_items_csv_file.html&scope=null    says
Planned For : The values of the Planned For attribute come from the project area iterations. Look up these identifiers in the properties dialogs of timelines and iterations on the project area editor's Overview tab. Replace each target ID with a path using forward slashes as separators, starting with a slash followed by the timeline's ID as the first segment, followed by segments of iteration IDs. For example: /main_dev/milestone_3/stable.
Tip: If a source value is not mapped to a target value, the import operation creates new iterations or timelines for those source values.

Seems any mapping for Planned For values as below, does not seem to be working as intended.

    <attribute sourceId="Planned For" targetId="com.ibm.team.workitem.attribute.target" name="Planned For">
        <value sourceId="/development/Test/Test Iteration 4" targetId="/development/Test/Test Iteration 4"/>
        <value sourceId="/development/Test" targetId="/development/Test"/>
        <value sourceId="*" targetId="" name="Unassigned"/>
    </attribute>

All the values used are identifiers.  Whenever value mappings as above are used for csv import,  "Planned For" get set as Unassigned in the imported work items.  CSV file has the value "/development/Test" for the "Planned For" attribute for few work items and  never gets translated and set even though  "/development/Test"  exist in the Target project area where the CSV import is done.

RTC 4.0.2 version used   for both server and client.

Is the value mapping not allowed for "Planned For" attribute ?   or Is this a defect  and any work around solution for this ?