I agree with your overall point, but I just want to point out that many (most?) journals don't employ copy-editors, or if they do, then they overlook many errors, especially in the methods section of papers.