Has anyone posted about the performance/reliability differences with the new home architecture?
I'd also like to know about this. My experience with 16.2 is that
(a) the new architecture is better WHEN IT WORKS. Unfortunately the number of situations in which it fails to work is very large (appears to include at least trying to share your home, perhaps having multiple homes, and definitely [my personal experience] adding a new Apple TV to your home).
The failure case is utterly catastrophic with every device in your home eventually losing contact with HomeKit and having to be "reset" (which, among others things, destroys all automations and scenes). To add insult to injury simply adding devices back does not fix things, they randomly over two or three weeks then lose the HomeKit connection again.
(b) the improvements in HomeKit 16.2 are with latency. They are not with anything automation related (either reliability or debugging assistance).
So basically 16.2 trades off a minor hassle (latency) for the (fairly high) chance at utter catastrophe, without fixing ANY of the real HomeKit issues. It's not a good tradeoff and the fact that the HomeKit team thought this was a remotely sensible upgrade just confirms that they truly are the worst software team in recent history, not just in Apple but across the entire world of tech.
The crazy thing is that Automations are THE ONLY THING where Apple provides some sort of differentiation or advantage over Ikea, AMZ, Google, etc. So of course the HomeKit team has utterly ignored this competitive advantage year after year!
If there are bonuses being withheld at Apple as the other MacRumors story says, uh, Tim, I have a suggestion as to where you might want to avoid that bonus money...