I tested a very similar device for the company I used to work at, the UI looked identical at least. There was very little relationship between my actual mental state, and the state of the display. The device I used had an API, so I wrote my own visualizer to make sure they weren't just blurring out the data -- long story short, I was never impressed with it. Blinking or moving your head triggered a larger reaction than any thought impulse.
Maybe the tech has gotten better, but I wouldn't hold your breath for anything that costs $100.
"Blinking or moving your head triggered a larger reaction than any thought impulse."
No surprise there, the muscles of the head will have a stronger signal than the brain, if you're using electrodes outside the skull.
There might be some way of filtering those out, perhaps using additional electrodes directly on the relevant muscles. Subtract the signal on those electrodes from the signal on the other electrodes.
Also, based on the geometry of the xwave headset it looks like the electrode(s) are very rostral, so I would imagine that it picks up a lot of EOG from the eyes and EMG from the frontalis and corrugator supercilii (eyebrow muscles).
The ground, again from what I can tell from the photo, is on the earlobe. This is not a terrible place for a ground (there are no muscles there) but it's not ideal for removing EMG artifacts.
Maybe the tech has gotten better, but I wouldn't hold your breath for anything that costs $100.