This.

Elsevier is not stupid. January without access means:

* most scientis realise it's not necessary

* the smaller rest of them, learn how to use scihub

* elsevier is 10% less on revenue (contrary to giving it 'free access' now and retrocharging later) for every month

* people start publicize in different places, diminishing elsevier value, and proving that you can print elsewhere (sic!)

In result elsevier strong grip ends in a bing bang. Would scientific world hold them for just 3..5 months, and they're out of the business. Simple.