Can this technology write to the thinking parts of the brain?
If the answer is "yes," then this can augment human thinking capabilities -- the rest is just improving bitrates of the human <-> computer channel and writing better software. Both of which are challenging and rewarding research areas, to be sure. But they sound a lot more like straightforward engineering with predictable improvements from research investment, rather than requiring basic research / fundamental discoveries.
Transcranial magnetic stimulation is a very blunt tool, and, in this case, the "mind reading" is really a binary switch that is triggered by thinking of something that produces a broad, easily detectable EEG pattern.
At best, repetitive TMS (rTMS) can bee used to temporarily make a brain area more or less active.
If the answer is "yes," then this can augment human thinking capabilities -- the rest is just improving bitrates of the human <-> computer channel and writing better software. Both of which are challenging and rewarding research areas, to be sure. But they sound a lot more like straightforward engineering with predictable improvements from research investment, rather than requiring basic research / fundamental discoveries.