From a particular point of view, the term client-side queuing makes absolute sense. The view is if the Broker is considered the “server” and IS is considered the “client.”
Thus, the IS ability to queue publications to the broker fits quite nicely into what is traditionally defined as client-side queueing.
The reverse, where IS has pulled a doc from the Broker and placed it in another queue for later delivery is a good thing too. But calling this client-side queueing seems strange. I’ve never heard this type of interaction referred to as such. Maybe I just lead a sheltered life. 
But my main concern with this terminology is that IS and the Broker shouldn’t really be positioned as one being a client to the other. I know that’s how things really are under the covers but I believe this leads to confusion. Developers think of all of the wM Platform as the “server” with the “clients” being the applications connected to wM.
For this POV, where apps are clients and entire wM suite is the server, wM doesn’t provide a client-side queueing facility. IMO, the term client-side queueing has been thrown in to address a short-coming when comparing with things like WebSphere MQ.
My 0.02. Opposing views encouraged.
#webMethods#Flow-and-Java-services#Integration-Server-and-ESB