Isn’t there also a blatant mistake in the content of the post? The naive approach for approximating e keeps track of two floats. No arrays. So minimal, aka neg

Isn’t there also a blatant mistake in the content of the post? The naive approach for approximating e keeps track of two floats. No arrays. So minimal, aka negligible memory footprint. Then they introduce iterators, promoting that yielding values is more efficient. Yet suddenly, using islice, they are actually producing and having to store a list of values. So now they do have to worry about memory, as opposed to before. Yet, the article claims/implies the opposite.