@danabw , WOW! As you were writing these things, I was experiencing them! One of my observations I'm becoming more and more confident in is there is some type of memory leak or operational stuck loop in the inclusion exclusion processes. I agree with you as well that it seems to perform better if I reboot after 4-5 inclusions. SO, definitely something awry there in the process. And, whatever is happening, I think can also be triggered by halting the inclusion/exclusion process. For example, if I start the inclusion and exclusion process after a fresh restart, and stop it because say I wasn't ready or had the device in the wrong state for an inclusion, and press the button again to stop the process, I will have more than likely re-created whatever is happening after 5-6 inclusions. Some sort of leak or stuck look that breaks down the whole of the z-wave infrastructure on the hub. No Z-wave related process behaves appropriately after whatever happens happens... Turning Lights on or off, running z-wave repair, nada. It all breaks down until a reboot...
I have not tried shutdown and power pulls. BUT, after reading your post here, and my experience, I am going to start trying frequent reboots as well as everything else you mention.
A final observation I have run into in the last 30 minutes (after trying z-wave repair because I was having so many problems), is that 1. The Route tables were ALL empty before z-wave repair... Three light switches and all 8 repeaters failed the z-wave repair process, and, after that process was complete, the route tables were filled in for most devices, but, missing on those 11 devices plus a handful of others. So, that was weird. to me. ALSO, the routing is UTTER chaos... Makes no sense. A device is running through another one 150 feet away, only to jump back to a device that's 5 feet from it. I truly don't understand how this routing is developed.
Is it normal I wonder for the routing tables to be blank after adding 30 items, prior to running a z-wave repair (originating with a fresh system)? And then, is it normal for all repeaters to fail the repair process?
These are my thoughts right now and yes indeed, I am in the storming and forming phase. But, for those of you migrating to HE, if it works like mine, indeed, frequent reboots/shutdowns while adding does seem to help. Hopefully once everything is added it will normalize/stabilize.