For the moment there isn't yet a plan to fix that.... It is usually an application flushing after closing so it should be fixable in the application it self.
The extra flush is often induced by an AJAX framework or a library, and the container-side fix, as easy as it is, may not always be within the reach of an application developer.
I think that an httpd-side fix would be more robust, and this can be relatively easily addressed by a change in mod_proxy_cluster.c:ajp_handle_cping_cpong() to detect and discard any flush message preceding a CPONG
Thank you,
--Alexey