Yes all those journals are using readcube or a derivative which is a nightmare. They just don't want people to see the pdf as a first option. On many journals you need at least 3 clicks and 2 loading pages to get a pdf (unless you use a citation manager like zotero that would catch the pdf from the article page).