It isn't clear here whether the papers in question were the pre- or post-editing versions
They are going after the final, published versions (including Elsevier formatting and all), commercial use of accepted manuscripts, systematic distribution and the like (some of which applies to academia.edu). In other words, what you said was fair game still is - you are allowed to share the accepted manuscript with others (including on your website where Google Scholar will pick it up and render it discoverable in a matter of days, so it's not like this restricts you), you (or anyone else) just can't make money off it and you can't use their typesetting.
For the accepted manuscript version, let me just quote from Elsevier's author rights:
Elsevier believes that individual authors should be able to distribute their AAMs for their personal voluntary needs and interests, e.g. posting to their websites or their institutionâ(TM)s repository, e-mailing to colleagues. However, our policies differ regarding the systematic aggregation or distribution of AAMs to ensure the sustainability of the journals to which AAMs are submitted. Therefore, deposit in, or posting to, subject-oriented or centralized repositories (such as PubMed Central), or institutional repositories with systematic posting mandates is permitted only under specific agreements between Elsevier and the repository, agency or institution, and only consistent with the publisherâ(TM)s policies concerning such repositories. Voluntary posting of AAMs in the arXiv subject repository is permitted.
So you can see how academia.edu falls foul of this while your right to share your work does not.
(Some of my papers are published in Elsevier journals - they are however also all open access. In case you're wondering.)