I am not sure if this is the main root of the problem between scientists and some commercial journal publishers, but it seems to be a problem between the non-profit altruistic mindset vs the profit motive mindset.
For the most part, scientists and scholars want to publish articles to gain scholarly recognition, academic rewards such as promotion and tenure, and for the prestige of publishing in a high impact factor journal because they only accept 10-15% of papers, etc. They are researching and writing for altruistic reasons such as advancing science, make the world a better place, helping others learn how the world works, etc.
Back in the 1950s and 1960s, scientists started using many more commercial publishers because there was a lot more research being done and they needed outlets since society presses were not able to absorb the large increase of article submissions. The commercial presses learned that libraries generally didn't cancel journal subscriptions because of institutional momentum, and an inelastic market for the articles existed. (Academic libraries keep subscribing because the patrons keep on demanding access. Patrons demand access even though they are not the ones paying for the content.) So, publishers such as Elsevier learned that scholars publish and review articles and edit the journals for next to nothing because they want recognition, not money. So, the publisher takes advantage of the situation, raises subscription rates because libraries want to please their patrons, so they don't cancel very often.
20-30-40 years ago, publishing information to hundreds or thousands of others was hard, printing and mailing was difficult, laying out articles was tough--particularly scientific content.
Then the Internet happened.
Scientists are smart and have figured out that global publishing is easier than commercial publishers let on. They are starting to take advantage of new OA outlets such as PLoS and others.
But, the culture of scientific publication had been ingrained in scientists over hundreds of years, and it takes a while for some scientists to change their perspective. They still want to get noticed, get scholarly rewards, get tenure, and get cited by others. So, they have lists of journals that junior scholars are supposed to publish in. However, some of the more forward thinking scientists understand that publishing in OA sources will increase their reach, increase their citations, and increase their standing in the academic community. (The Wellcome Trust says
“it is the intrinsic merit of the work, and not the title of the journal in which an author’s work is published, that should be
considered...") They do not need to submit their articles to closed-access (toll-access) journals. There is a reason that PLoS One is the largest journal in the world with a solid impact.
The profit motive of some scientific publishers has skewed their perspective. They do not really care about maximizing the reach of scientific knowledge. They want to hide the articles and the research, because they are used to an age of information scarcity, but we no longer live in that time. They are trying their best to hold on, and it shows. They may claim that they support universal access or something like that with a sustainable business model, but that is code for--"We want to keep on sucking more and more money from libraries and subscribers for as long as possible. We like our 36% and 42% profit margins, thank you very much. We will provide access to a small segment of the developing world because it makes us look good, and we don't get much of an increase to our server load--all we have to do is turn on some software switches to some IP addresses."
I know that there are other variables and opinions, but this seems to be the crux of the problem. Most scientists don't care about making a profit off the research, and they want to be cited and read by as many people as possible. The commercial publishers that used to be seen as supporting the scholarly communication interests of scholars have diverged too much. They want to lock away research and knowledge. The two groups are diverging, and the scientists no longer need to support the aims of the for-profit commercial publishers.