> it claims all parallelization methods are the same, so the only dimension to choose on is memory efficiency
From a speed (overall script duration), they are very similar and the differences between them might be as well caused by a different state of a network. My main interest was how the methods perform regarding memory usage. That's why I focused on one dimension only.
And it's fair to care about memory usage, but I think it's unlikely that your circumstances regarding runtime performance (both your testing script and testing environment) will generalize to other scenarios.
