JSTOR has happily allows publishers to restrict documents which are lawfully in the public domain and over which the publishers have no say except via the TOS they clipwrap you with when you try to access them.
Excusing JSTOR is like saying "Hate the game, not the players". While JSTOR isn't doing everything in their power to fix this broken system I say: Hate the game _and_ the players.
Ignoring the fact public domain gets complicated when you're an international organization (there's many journals which are public domain in the US but aren't anywhere in the world because of historical anomalies in US copyright law), JSTOR is essentially built on relationships.
JSTOR has access to the current archives of the journals because of the relationships it has with the publishers, if the publishers weren't happy with what JSTOR were doing they could just pull their licences for their current journals. If that happened JSTOR would basically collapse.
It just doesn't make sense for JSTOR to do it, when a third-party who doesn't have any relationships to the publishers could do it just as well (you can easily get physical copies of most historical public domain journals either via the open market or via libraries and scan them).
JSTOR does not have the power to direct what others do. If the publisher chooses to restrict a document, JSTOR has no say in that decision--JSTOR is merely a repository. JSTOR has access to essentially every academic article in existence precisely because it does not attempt to enforce any political goal on contributors.
If you want to change what publishers do with articles which should be in the public domain, get the law changed. But don't hold JSTOR responsible for something outside of their control.
The law already is what it needs to be. JSTOR has limited access— and continues to do so through their terms of services— access to even old documents which are already lawfully in the public domain.
They have operating costs. They're a non-profit but they pay people to digitize the works. It costs a lot of money to access some of these things. It's not prohibitively expensive to access these documents. You can read three articles free every two weeks if you're exceedingly down on your luck.
You couldn't a week ago. This isn't some accident. Their openness comes hard won with threats of their reputation.
Their digitization is largely done by a for-profit company. It's not competitively bid. There are many parties who would digitize these works at no cost— archive.org and google for example, and could afford to do it without pay walling the results.
Anyone at all (including you) is free to scan journals that are in the public domain and publish them online for free. If you scan/ocr them then there are many places (including Project Gutenberg and archive.org) who will host them for you.
Excusing JSTOR is like saying "Hate the game, not the players". While JSTOR isn't doing everything in their power to fix this broken system I say: Hate the game _and_ the players.