Agreed, especially if you use custom lua in scene-triggers and/or startup.lua. Beyond that, perhaps some sort of scene migration tool would help but device identity changes make that difficult too without a manual mapping. As an aside, I’ve been moving away from custom lua code, scenes and various plugins to Reactor because of the clear benefits but also quietly hoping that doing so might make the transition easier - @rigpapa is going to have to transition at some point too so maybe he will formulate some Reactor-based migration ideas.
It is early so I haven’t given it a lot of thought, but maybe the ezlo controller grouping could be extended, if it isn’t too much work or impossible, to include Vera’s sufficiently to allow an ezlo controller to orchestrate legacy Veras (i.e. bridge) until users are able to migrate various automation domains to ezlo. I imagine that any migration investment will have to be weighed against the benefit because it is a one-time penalty.
That’s pretty much where I am as well and have posted as much in the past. I look at every shiny new option that comes along, often buying them for experimentation but the original Vera has stood the test of time and this community is the best of the best. IMHO, if you add up everything the ezlo guys are doing across their cloud services and controller/hub, even the Vera will pale in comparison, let alone the current competition. There’s still a long road ahead, but it the path leads to a bright future.