The low quality problem with primary research publications is not the writing but poor experimental design, misrepresenting experimental results, shoddy statist

The low quality problem with primary research publications is not the writing but poor experimental design, misrepresenting experimental results, shoddy statistical analysis, and putting null results into file cabinets. Summarizing research findings isn't the crux of the problem, so if anything if an LLM can help the author perform a clearer and more concise writeup I'd see it as a net benefit.