They hybrid publishing model was adopted by the big scholarly publishers as an interim measure, a rearguard reaction, to the demand from researchers, funders and librarians for Open Access. It works like this. The journal stays behind a paywall, available as part of a package or journals (no-one buys individual journals subscriptions anymore). If the researcher wants (or is required) to have their article available as Open Access, they pay an Article Processing Charge. This is to reimburse the publisher for any lost revenue from lost subscriptions, and also places a fig leaf of progressive behaviour on the publisher. This is how Elsevier can claim to be “a leading open access publisher” 
From a library perspective this is tricky. It means that a proportion of the articles that are arriving as part of a subscription bundle are actually Open Access. As an institution we need to come up with the fee to pay for Open Access for a researcher, and still buy the journal subscription. This is called double dipping. Some of the smaller publishers (I’m looking at you IEEE) have worked really hard to resolve this transparently.
Why are we in this position? When the UK decided to support mandatory OA they plumped for what got called the ‘gold’ model in the Finch Report  This was a half-arsed position that encouraged a model that siphoned money straight from the publishing budgets of Universities straight into the publishers.
Some five years later, and the culture of paying for fees for OA has lead to low impact publishers creating a ‘pay to publish’ business mode, with little if any quality control. This with the swingeing  APC and subscription costs from the ‘quality’ publishers amounts to what we like to call ‘predatory publishing’.
So, what the fix? Two big things (and one little one) have happened to combat this OA mis-turn on the part of the UK, and copied in the rest of the academic world. Wellcome – a big medical and science funder – is not letting its funds recipients use hybrid . Plan S has softened (ever so slightly) towards hybrid, but only if publishers are using it as a way to permanently transition to full OA. . The University of Canterbury don’t support applications to their Open Access Fund for hybrid publications. 
This means that funders will increasingly turn to not supporting the predatory hybrid journals, where double dipping for Article Processing Charges and subscriptions is rife, and encourage journals to flip or be flipped.
- Even Elsevier state this is a “surprising fact“
- Thank you Manchester Metropolitan University – a definition of the Finch Report