Definitively a useless, if not unfair test...
We don't have a clue regarding the topology of the tested setup. Powerline indeed presents some connection imitation, typically if the modules have to connect through a circuit breaker, or if power wiring length is important (which can be the case in some building toplogies, even if the two outlets are in nearby or even the same rooms). The same way, test Wifi bandwidth beyond a steel wall and you'll barely get a connection at all. Besides, a Wifi transmission can even get impaired by a body going walking through the line of sights of the two Wifi transmitters.
If you do want reliable and efficient network connection, there's no alternative to Ethernet for now. Any other "no-new wire" technologies (HomePlug, Wifi, Moca, HomePNA, etc) have all their pros and cons and should be considered as complementary. Due to heavy cable wiring in the US Moca is an interesting solution in this market, but in Europe typically, Powerline (being it HomePlug or DS2) is widely used for in-home IPTV distribution, much more than Wifi is.
Any serious test would consider several topologies, not only in lab but also in typical homes. Besides, I don't know about the Intel NASPT tool, but I'm a little bit puzzled by overall results, especially by the figures provided for the video streaming in Wifi.
I suggest the tester to reconsider more seriously their case. You can easily find a situation when the test can provide exactly opposite results, demonstrating weakness of Wifi and extraordinary PLC figures (a situation which would neither be representative of on-the-field behavior).
I definitively would have expected more professionalism from a site as Tom's Hardware.