I do not agree with your conclusions based on the two recent review papers by Simons et al., (2016) and Melby-Lervåg et al., (2016, p. 526), both coming to the conclusion: “… that there is no evidence that working memory training yields improvements in so-called far-transfer abilities”. The meta- analysis are to some extent biased excluding studies that would substantially contribute to an opposite viewpoint. Melby-Lervåg by no obvious reason excluded the COGITO study (Schmiedek, et al., 2010). In this study more than 200 individuals were assigned to the training group. The number of participants contributes most to the estimated power a study adds to a meta-analysis. The study found evidence for far transfer effects of the training which was present also 2 years after the training ended.
Similar, Simons, et al., (2016; p., 111) in support to their claim that WM training shows only near transfer effects, cite studies on neural plasticity, for instance, that violinists displayed selective neural growth in the right motor cortex, corresponding to the use of their left hand to finger the strings. On the other hand, they fail to notice the Owen, et al, (2005) extensive review of brain related changes following n-back training. The review provided robust evidence for the activation of 7 brain areas in all n-back variants showing strong evidence for a relationship between n-back practice and WM. The findings were replicated by Kearney-Ramos et al. (2014). Similar findings were reported by Román et. al., (2016) and Thompson et al., (2016). Moreover, the identified changes in brain areas following WM training also correspond with those being crucial for intelligent behavior identified by the two most influential theories of the relationship between brain structure/function and intelligence – the P-FIT (Jung and Haier, 2007) and the multiple-demand system (Duncan, 2010).
We discuss this issue in our blog:increasingintelligence.blogspot.si - Increasing Intelligence