Polyglot has made some good points...especially the final part about "huge fun". They're definitely fun. As I said early on, these are at once simple and under the surface more complex. If not abused, they will work for a long, long time. I would personally suggest that a person design a machine or appliance which would be an occasional use thing...that being something you need to turn on to accomplish a chore and then turn off when no longer needed. Efficiency can be a slippery thing when you step into the world of refrigeration, and common sense is a good yardstick to remember. Example: I showed you how I created ice in 5 minutes. This also included the burden of 80 degree ambient air and cooling about 4 cubic inches of aluminum to a temperature below freezing. Now, let's consider this: the ambient temperature inside the freezer compartment of your refrigerator is ideally 0 degrees (f) but more realistically it varies between around 10 to 18 degrees (f) ... in a typical appliance, that is. Mine hovers around 14 degrees most of the time. How long will it take to make ice if you place the same quantity of water I froze in 5 minutes in an aluminum cup in there and then set that aluminum cup on top of another piece of aluminum which is at a temperature of 80 degrees (f)? Don't worry about going to try it...I can answer that question right now. It took me over 25 minutes. How about if we forget the 1" thick aluminum spacer? How about if we just put the same quantity of water in the aluminum cup and set it in there? With the fan running (which will improve efficiency of your freezer obviously), it took me almost 15 minutes to get the ice as firm as it was on the test yesterday. So, if it takes you between three and five times longer to freeze that quantity of water using a more traditional appliance, which approach is more efficient? I'm not arguing...I'm just adding the idea that factors other than a simple measure of wattage should be included. You know, often it will take one man all day to do what 5 men can do in a couple of hours.
The "thermoelectric effect" is really 3 separate effects. The Seebeck effect which has to do with temperature differences converting into electricity, the Peltier effect (which we've been discussing to this point)...the heat transfer and all and the Thomson effect which as I recall wasn't discovered by Thomson but Kelvin. It has more to do with coefficients and is more complex than I care to wander into this early in the morning.
But yes...you definitely can generate electricity by moving heat through the device and waste heat can be put to good use that way. Thanks for pointing that out, Polyglot.