With the wealth of data that they collect, this would be a great place for an economist (like myself:)) to study usage patterns and possibly conduct experiments.

Of course, the set of people that buy this device are more "green" than the average consumer and also more likely to change habits. Nonetheless, looking at correlations of household-level usage with location specific variables like weather, sporting events,etc. could be really interesting. Moreover, conducting experiments where alerts are different for 50% of the users to test how responses differ would really illustrate how public policy should move forward. Think of the A/B testing web developers use, but with possibly wide-ranging consequences.

Great work!