Yes you are correct of course. ERA5 is climate model output constrained to match weather observations, not actual observations, which I do note in the post.