You are making the wrong assumption that effect/no effect are equal probability of being true.
However, the way science works is that you have to prove an affect. The null hypothesis is that brain training doesn't have any affect on "performance". To assert otherwise, you need evidence.
>>Neither the above quote nor the debunking of their internal studies as junk equals any proof to the contrary, that brain training -- or this product for that matter -- doesn't work.
The article isn't meant to be proof that it doesn't work. He is saying that there is no good evidence to believe it does.