I think that's the problem: You've *just* got the imp. By itself, it doesn't do anything. The point is that other devices will be built *around* the imp, and *those* are what will be cool. The imp fills a technological gap for any inventor who can build a neat physical device, but can't make good web-enabled software.

There are tons of devices that would benefit from being web-enabled, if only the people who made those devices could write decent software. There was always a lot of talk about web-enabled appliances, but it turns out most of the appliance manufacturers are only good at making reliable appliances. An imp could, for example, text me when the dryer cycle was done so that I could run upstairs and pull out my dress shirts before they got wrinkled (I'm usually out of earshot of the buzzer). But the imp doesn't do that by itself, it needs the circuit to integrate it with the dryer. You need both bits before it's useful.

It's like the Arduino. You could buy an Arduino Uno board and stare at it and say "well this is useless". Or you could design a circuit that does something cool with it. Or, you could buy any number of devices that are already made by someone else, which happen to *leverage* the Arduino internally.