Part of the problem in teasing out the effect is that there is an inherent selection bias in the type of people who receive this type of education (or any education, in the not too distant past) to conclude from population analyses to what degree the particular task of literary memorization confers this benefit. Someone's going to have to ferret out the mechanism or do controlled longitudinal studies, which will be costly and fraught with difficulty. Even Socrates (i.e. Plato) was attuned to the idea that memory was under attack with the advent of writing[0] and a culture that has other information retrieval technology, so it's debatable whether we could really plumb the depths of what human cognition could be under different pressures. It's also unclear how early this kind of training has to be started too in order to have an effect.
[0]https://blogs.scientificamerican.com/literally-psyched/on-wr...
I do find it plausible that certain (combinations of) language and verse metre might have mantra-like qualities, conducive to achieving trance states that could affect the brain. India has been on the case for a looong time, so it wouldn't surprise if Sanskrit + Vedic poetry present a particularly sophisticated example.