Handling a comms message has side effects #432

When handling comms messages, the kernel calls shell.events.trigger('pre_execute') and self.shell.callbacks.fire('post_execute'). This limits the utility of comms messages as sending a comms message to the kernel has quite substantial side effects. For example, any pending matplotlib figure are plotted as soon as a comms message is received. I am using comms in spyder to communicate between the kernel and the frontend. Is there a reason why pre/post_execute need to be triggered when handling a generic comms message?
I am using monkey-patching to remove these calls in spyder, and I was wondering if this is something that should be fixed upstream.

Line for ref: