You sound like the slashdot review of the iPod: even if all of the features were possible before does not mean there's no benefit to producing a solid, high-quality implementation.
As for Nest's claims, I assume you're talking about energy savings and, if so, are so very, very wrong: I'm a software developer and rarely adjusted the program on my old Honeywell thermostat because it took a minimum of 7 * 3 button clicks (morning, evening, night for each day of the week) simply to navigate through the program. I'm rather confident that the average American is even less likely to put up with all of that clicking and simply leaves at anything which isn't uncomfortable, even if it does waste power.
Not really, unless your world is binary. There are a lot of extremely refined, slick implementations out there. You don't know about them because they couldn't be called the "ipod of thermostats" (with leading comparisons with the ugliest, most rudimentary thermostats, as if the giant industry doesn't exist).
As for Nest's claims, I assume you're talking about energy savings and, if so, are so very, very wrong...I'm rather confident that the average American...
It is interesting how you arrived at such an energy claim with no clear avenue between the beginning and the end.
You asserted that it was “a non-solution for a non-problem” - the onus is on you to explain why the stated problems aren't real or how the Nest doesn't solve them.
As for the promise of saving energy, either reading Nest's citations or spending a second or two on Google might prove educational as to the current gap between what is technically possible and what people actually do: