I just responded to this in the Brain-Computer forum, but it's worth repeating myself here.
A good replica of the test,
http://cognitivefun.net/test/8
Or if you prefer German phonetic,
http://cognitivefun.net/test/5
I agree that this looks pretty important and significant. I've started two days ago and gone from being completely lost on the 2-back to doing fairly well on the 3-back in total play time of no more than 40 minutes. By the end of the month I'll try to make some assessment as to how this translates to my everyday function. Subjective and anecdotal, but if the results are as significant as claimed, I should notice an improvement in my routine.
that's handy.. I have written my own multi-platform version of it based on the actual study (same basic stimuli, essentially the same rule the task uses to make it harder or easier depending on your ability, etc.).. just wrote it for fun as a way to learn Qt
I do think an important aspect of their study was the way the difficulty of the task was constantly being adjusted.. also, the task gets really hard: 4-back will make you suffer!
actually, quick look at that page suggests they too based that on the paper originally mentioned, so that's good
more interesting is the question of how long the effect will last.. a colleague of mine and I chatted about the study briefly and she wasn't particularly convinced it would be useful for increasing intelligence, whereas I'm not so sure.. she, however, actually does this sort of testing on healthy and diseased humans, whereas I only do this sort of testing on software, so she may have better intuition