Document how the large message support for Core is different from other clients

In the "13. Working with Large Messages" chapter, mention how the large message support for Core is unique from other clients"

Following is the comment from Justin Bertram [1]:
The large message support for Core is unique because it is the only use-case where the client itself has special support for large messages (i.e. it can stream "chunks" of the message to avoid memory issues). In all the other protocol use-cases the large message support is strictly on the broker to avoid memory issues when dealing with large internal buffers.

[1] https://issues.jboss.org/browse/AMQDOC-2663?focusedCommentId=13557406&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13557406