Sure, the energy equation for extracting hydrogen isn't awesome (though I suspect if you ACTUALLY boil down ALL the inefficiencies in the electric-car-based-on-LiIon-batteries equation, the actual "joules in to miles traveled" ratio likely favors hydrogen by a long shot). But who cares if the source is, for example, solar, wave, hydro or wind energy from a station close to the sea, which also happens to be a great source of non-potable (therefore not competing with human drinking needs) water?
The average age of a vehicle on US roads is 11.4 years ish and climbing. Self driving cars, like home automation, are "five years away from changing the world" and likely always will be; definitely still will be in five years. Minor aspects of functionality originally developed for self-driving applications will become mainstream piecemeal, but we're decade(s) away from self-driving cars being mainstream.
Because environmental activists are trying to ban THOSE sorts of bulbs.
They may never issue another firmware upgrade for these particular hubs; simply, the next version of the hub will be marketed as "for Friends of Our Wallet Certified Partners Only" and will be incompatible to non-partner devices from the get-go. It is absolutely conceivable that this was truly a UX decision - trying to tamp down the level of complaints from consumers who bought third-party bulbs that don't quite work right. However the fact is that this is a nascent (many might say, unnecessary luxury) market and people who buy this stuff are almost exclusively bleeding edge technology buffs and tinkerers, or people who simply throw a blank check at an integrator and say "make it work". The latter category of people isn't generating these support calls, because their integrators buy the expensive bulbs to avoid tech support, and the former category - which is the enthusiast category that could grow these devices into the mainstream - demands interop.
The reason computer chips are so small is computers don't eat much.