The wind is blowing towards gold open access, i.e., the research remains published at Elsevier, it is open-access, but publishing each article costs $1000-$2000 charged to the university via grant money or "subscriptions to publish" in place of the existing subscriptions to read.

I don't think Elsevier is really worried. This gold OA model is just sending the same quantities of money to them, just under a different name, and with the side benefit that everyone can read the published articles.

The real change will be when researchers stop publishing in the journals that they own -- noting that the costs to publish are so high because the prestige of journals means that fair competition can't happen. Sadly, it doesn't look like researchers are willing to take such a step -- or even to stop reviewing for these journals for free (see our pledge in this sense, https://nofreeviewnoreview.org/)