I'm a big user of Micasaverde systems. I've been installing them as a security professional for a year or so. I agree, the UI is very bad. I don't agree that it's overly complex, however. I think that versatility (even if clunky) is the entire value of the system for a guy like me. Without those off the wall features, it isn't a true HA system. It's got to have as many options available to it as it can.
One thing that Micasaverde lacks is a proper control keypad. I've toyed with rigging Tasker and Nexus7s with Paranoid Android... but it's time consuming, costly, and can be a service nightmare for a businessman who enjoys reliable installations. Almond+ would solve some of these issues... but I'm getting the impression that they are going to take the same tact as Micasaverde did.
It's tough to be everything to everyone. Especially with a small company, trying valiantly to get their product to market on time. It's hard enough to satisfy production demand as well as trying to incorporate support for the five-thousand-three-hundred-forty-two different devices that their users are requesting. The only viable solution is to do exactly as Micasaverde and other HA platforms did; Support the community itself, and aid users develop their own drivers for this hardware.
If Almond+ sticks to just Z-wave and Zigbee support, it's not much different than something like the upcoming Piper. If it can manage to get enough drivers for other systems in a timely manner upon launch, it has a great hope to be a stellar alternative to established HA companies. I'd switch, and I'd put in dozens of them.
They key approach that differs from companies like Micasaverde, is that their success in integration occurred by accident. Promote the developers, and look to the professionals to guide them on what works, what doesn't, and how to go about it. Micasaverde isn't even aware guys like me exist, yet I've purchased a dozen or so of their units, for my clients in 2013 alone. When counting my colleagues and competitors in the industry, we've probably put in a hundred or so of them last year. That's likely a large chunk of all sales of home automation systems in my small city. The infuriation I feel when I see a killer feature isn't fully implemented is palpable. All because the end user who wrote the driver for free got bored with that project, and never updated it.
Sorry, I'm ranting. A bit off topic as well. Forgive me.