AAM - global acceleration for web application - how does request get to data center big-ip?

Straight from the guide:

About symmetric request and response headers
In a global network that includes a symmetric deployment of remote and central BIG-IP®
devices across a
WAN, the remote BIG-IP device receives a request and includes an X-Client-WA header, which
distinguishes the request to the central BIG-IP device, enabling the central BIG-IP device to process the
request, as necessary. When the central BIG-IP device receives a response for the origin web servers, it
includes an X-WA-Surrogate header in the response, which distinguishesthe response to the remote BIG-IP
device, which processesthe response as necessary and removesthe X-WA-Surrogate header before sending
the response to the client

If this is the case then I m not quite sure why we are configuring a default pool for the remote big-ip. I have to believe its in case the central(data center) big-ip goes down?