I recently bought a UPM EM-100 "electronic energy meter" ($25 at Canadian Tire). It measures line voltage and current drawn by whatever is plugged into it, calculates "watts", and records max current, max "watts", operating time, and the integral of watts. It will also convert watts to dollars if you tell it your energy rate. It looks like this:
Internally, it just uses a shunt to measure current; there's no current transformer. The circuit is battery-powered and connected directly to the line. Current range is 0-15 A with 0.01 A displayed resolution. I compared it to my best handheld meter connected in series with the load, and the EM-100 reading agreed within a percent or two of the "serious" meter.
However, it seems that it really measures volt-amps and calls that "watts". I plugged in an unloaded isolation transformer, which would have had a very low power factor (almost pure inductive), and the measured current was quite accurate, but the displayed "watts" was simply the product of current and voltage. So this unit should work well for high power factor loads, but it will be quite inaccurate for low power factor loads.
The most commonly recommended device for measuring appliance energy consumption seems to be the "Kill-A-Watt". It does distinguish between watts and VA in display, and can also display power factor. But is it accurate for low power factor loads?
Dave