Brain training doesn't usually transfer. The Posit studies haven't been much better than any others.

Okay, I played most of the free exercises, and apparently I'm like the ultimate boss at spotting different birds (aka "Brain speed - Hawk eye"), never making a single mistake at even the highest available speed, and merely mediocre/slightly above average at other things. I also noticed while playing the object tracking one that what allowed me to do better is I came up with new "algorithms" for tracking things. First time I did it, I tracked up to three objects easily, but then failed miserably at more. After practice, I learned to imagine lines between the objects, which let me track four correctly most of the time, and five occasionally. Which, setting aside me not be that good at this one, seems like a case of other-optimising. I really doubt learning to imagine lines between objects generalises well.

So, from personal experience, I'm sceptical it's useful, but at the same time, listening to the video in the background (which may have reduced my performance on some of these), it does sound like there's some research to support this.