I purchased a "hfield Wi-Fire" adapter recently because it claimed extended range. I found it new in the box at a discount store. It's a 2.4 ghz only adapter. It comes with a small clamp to allow you to attach it to the top of your laptop. Otherwise it is pretty standard, connecting to your laptop via USB.
When I plug it in, I can do the following on my Ubuntu desktop to get the transmit power information.
ericu@ericu-desktop:~/Downloads$ iwlist wlan0 txpower
wlan0 unknown transmit-power information.
Current Tx-Power=27 dBm (501 mW)
I was intrigued by this, because 501 milliwatts is a huge transmit power compared to most devices. I am pretty sure this number is just reported by the driver in the linux kernel, so it could be programmed to say anything. I opened up the unit, which was quite easy.
The design of the internals appears to be a typical WiFi USB stick feeding a yagi antenna. The drive element appears to be a partially folded dipole. I'm really unsure what functions as a ground plane in this design, unless it is the coax cable itself.
There is also a patent number listed which is 6307524. More information on that patent here. The patent appears to cover the PCB design for the antenna only. The description is a little curious, since it describes an antenna with the transmission line impedance matching the driven element. This is a desirable quality in most antennas. I guess they are really just saying this antenna design does not require a matching transformer.
Is it possible that this WiFi adapter actually transmits 501 milliwatts of power? Or does that figure include the effective gain of the Yagi antenna printed on the circuit board?