Another solution is to work backwards. Being in the security industry, I can suggest that a vast majority of old alarm systems out there can cheaply or easily be refurbished to work with home automation systems of various sorts. Micasaverde, Homeseer, Fibaro, Zipato, etc are all gateways that can accomplish this. Hopefully the Almond+ will help people in this regard as well, giving more choice in the market.
I say this because the security industry does not suffer from the lack of volumes. As a result, these sensors are cheap. They also are very well refined by this point, offering extremely reliable components. Batteries in most wireless sensors may last the better part of a decade in some cases. The same can not be said for Z-wave or similar. In those situations, the sensors are over priced, known to be less than perfectly reliable, and offer other pitfalls.. For example, the motion detectors are pitiful. Zero pet immunity is unusable for many people.
As for actuators, such as what you're specifically speaking about, quality can only improve as manufacturers get feedback as to how their products fall down. It's everyone's collective responsibility to let them know how their devices are being used, and how to improve their reliability. Costs can only go down once volumes ramp up.
Apple and Google will both make a play to enter the market with some devices. I wouldn't be surprised if even Blackberry gets into the gateway and cloud service automation business as well. (Whatever you might say about their company's failures lately, their network security is a thing of beauty) The issue at hand is to build more affordable, attractive and reliable actuators (thermostats, garage door opener widgets, blinds motors, etc). It's also to ensure that they can be compatible with all the various gateways out there. (Published APIs or SDKs, a willing community of coders able to port these new devices to the various gateways such as Almond+.
The market is in such a state of flux that I figure it won't really settle into any form of equilibrium for some time. Perhaps a year or two. Then you'll see things take off quite quickly, pushed by the big data players looking to acquire more data to sell to advertisers, or to just generally open up more markets for themselves. I look on it as potentially the next big thing, similar in scope to how the smart phone market changed society. Big things await us, I suspect. All the DIYers that have been waiting for it will simultaneously rejoice as their world opens up, but also may not entirely like where the big corporations take their hobbies...
My 2 cents. (Or was that 10?)