I have seen issues both ways, either from chatty devices just sending lots of commands.
Also try putting all your zwave devices in one group with metering OFF and then turn it on and off. What I have seen is this basically kills the zwave forcing a hub shut down to fix it. I had shared a large group to Alexa and then one of her hunches at like 2AM, the group was "on" due to one light, so she tried to turn the entire group off. Which she flashes lights 3 times before turning off to alert you. Lets just say I woke up and nothing worked, and half the lights were on from her attempts at flashing the group. Took me a few days of this happening over the course of a week before I figured it out. I possibly even had metering turned on in this case but it was too much.
You are SO lucky my wife wasn't sleeping at your house those nights. When I first put in some flashing light automations for leaks and stuff, I accidently left a test light-flashing automation enabled, and somehow had set it for 2 AM instead of 2 PM. That did not help w/my popularity at all.
I have had issues w/some Z-Wave devices making frequent route changes and eventually losing routing. Happened on both of my C7 hubs. I'm thinking it might be FLiRS related, as the issue seems to "follow" my battery-operated iBlinds. They were losing routing on my main C7 off and on (though sometimes continued to work normally). Moved the blinds to my other C7 (which was virtually empty) along w/a couple repeaters, and I've seen the same issue occurring there now. I'm hoping that the Z-Wave update will provide some help w/this issue.
ok, i want to timidly wade into this thread. I have had z-wave issues for a long time. right now, no ghosts, almost all devices direct connected, no spamming (near as I can tell), i even did the antenna mod... I have devices that simply just don't want to work when sending commands either directly or through routines. i've removed and repaired devices that appeared to be problems, let everything sit for days, run individual repairs, shut down the hub, you name it, i've tried it...
then i find this thread a week ago and start following it... i seem to have started encountering more issues, the more z-wave devices i have added in the last year... could i have impacted my mesh by adding to it? is it too big?
right now i have 110 devices in my network. all are plus and like mentioned earlier, most are directly connected to the hub... could i have the lockup issues @JasonJoel mentioned? is my only hope a fix in the stack like discussed here? i hate to bother everyone, especially those of you that are way more advanced than i... i'm just looking for some assurance that i have indeed tried everything... maybe we should be talking about this update more...
maybe this is the root of a lot of the s-wave problem threads i see pop up almost daily? i get you don't want to send everyone down a rabbit hole... i also get people have ghosts, etc., but i wonder (maybe not fully understanding) if a lot of issues are related to this?
There are also inadvertent situations that cause issues. The other day Alexa heard "turn on all the lights" when I said "turn on the hall lights". Mesh had finally started to somewhat heal after my power outage last week. That one misunderstanding by Alexa that turned on 50 devices at once killed the mesh again. Not really a situation where I could add delays even if I wanted to. Delays help in some situations I can control, but unless the delays are implemented at the hub firmware level we can't prevent floods entirely.
tbh, I am strongly inclined to believe this IS at the crux of the major issues.
HE was one of (the?) first 700 series hub. Thus, folks like me were among the first to start putting them to the test.
There have been issues, grumbles, complaints for some time.
The HE folks have really worked on fixing S2 issues regarding protocol--and they even added the metering feature to "Scenes" to spread the traffic out. And, those certainly helped...
But, there are limits to what they can do to work around a basic flaw in the protocol.
Based on my issues, I've long thought (and commented that) it seems that the hub was going "radio silent" at times.
In my mind, some of the remaining issues I have and have seen others talk about could all be caused by that...
Crazy routes going all over the building - If the mesh flips out and end devices cannot get ACKs from the hub, those end devices will start panicking and will try all manner of routes to get to the hub. With Z-Wave Plus, they will likely settle on the last one--which, in this case, is some crazy, out of the way route that happened to be the one that worked when the hub woke up from its coma.
Things like sirens repeating sounds...if the hub doesn't process an ACK back from something like a siren, it may try to send that command over & over. Thus, sending 1 "beep" could end up causing several beeps once the queued up commands on the hub actually get sent out.
Obviously, if the hub can't send things--all the things it's trying to do will be delayed (whether because they are queued and finally sent--or because it retries after it wakes up).
Scenes where many devices simply don't respond. Without metering, this could certainly create a busy mesh that triggers the issue.
Issuing just a very small number of commands works pretty reliably--but more than 5-10 rapidly becomes less-so.
As I said-it seems the HE folks have done all they can on their end--so we're all waiting on the SiLabs update(s) to hopefully help this out. And, understand, that this is a "first pass" by SiLabs at a fix. It's possible that it may take future updates--and this update may still have more and worse bugs that render it unfit for use.
I'm trusting in HE's testing, beta testing, etc. to help sort that out--then, once a properly vetted firmware update is ready, I'm sure we'll be seeing it.
The HE folks have def jumped in to show that they are on top of this--and they have more to gain that any of us individually. And
Don't think it's in beta release yet, but evidently there is a 1.17.2 release, so SilLabs does appear to be chasing issues w/new builds. Heard about it from a Z-Wave HW vendor.
I mean, in general, it seems there would be. But, also, it seems soon.
Did you get any clue what the issues were?
Following the links and nerding out, it seems there are still some issues. Of main concern, some folks are talking about devices "dropping out". But, it seems a "ping" brings them back online (often permanently). And, there are some issues with "healing" (that may relate to battery devices).
I finally saw a hint about the original problem. It seems the 700 series added some sort of "network check" where the hub refuses to send anything if it thinks the network is rather busy. So, it can sit a long time (not even sending ACKs--which causes end devices to start flipping out and resending things, making it worse).
Apparently, SiLabs is doing some tweaking on this. Tbh, I now wonder how much they can do without requiring EVERY device to be updated (not a likely thing, as a number of popular vendors don't offer Firmware updates).
I think this will be true and workable for most people but it really isn’t acceptable for the underlying protocol your home depends upon or I hope for Sigma ..
When I first put in some flashing light automations for leaks and stuff, I accidently left a test light-flashing automation enabled, and somehow had set it for 2 AM instead of 2 PM. That did not help w/my popularity at all. 
And 
really hope we see this update soon. Zwave will just “lock up” and stop reporting/responding.
