GzipFile.readinto() does not utilize the provided buffer but instead allocates a duplicate buffer and then copies the data at the end, which is the behavior of the io.BufferedIOBase.readinto() default implementation.
I have a demonstration of the problem in GzipFile.readinto reads full file before copying into the provided buffer · Issue #128646 · python/cpython · GitHub and I’ve submitted a fix in gh-128646: Implement GzipFile.readinto() functions by effigies · Pull Request #128647 · python/cpython · GitHub. I believe the fix is very straightforward to review.
Following the advice in Lifecycle of a pull request, I pinged the issue a week ago and am now posting here.