> That is of course an easy solution but I would argue that this is just throwing more resources at the problem. Not a very impressive optimization.

You could say the same about the existing implementation as that reads the whole file into memory instead of processing it in chunks.