My experience of the ACM was of paying for a year's worth of the Digital Library and finding that _everything_ I searched for in my area required an additional fee. It was listed as available but when you tried to download it would hit you up for another payment. When I complained about the bait-and-switch and requested a refund they refused.
The ACM just seem really underhand in general. They kept spamming me, through several email accounts, offering me membership. Even after I joined, they still continue to spam me.
I don't think I'm going to renew my membership when it comes up.
They also stepped up spamming me from quarterly to bi-weekly. I asked them to stop - no change - so I filter out everything from them. Now they just don't exist to me.
I'm not longer a member of the ACM because of their position on this issue. If you currently pay dues, take a moment to reflect on your reasons for doing so.
FWIW, it's often possible to find the full-text articles on CiteSeer (http://citeseer.ist.psu.edu/) or the original author's home or faculty pages via Google Scholar.
Also, while I graduated several years ago, I still have access to surprisingly many research journals via my alumni library card. If you are or have been a student, that's worth looking into.
Why I don't really care: most of the people who benefit from academic research papers fall into one of the following categories:
1. Academic professionals.
2. Students.
3. Industry professionals.
4. Government professionals.
In all three cases, there is an entity that will gladly pay the paltry sum required to get you access to ACM papers should you have even the slightest realistic need for them. You have to work pretty hard to be in a category where the dues will be both a significant expense and not paid for by someone else.
That is, it's quite silly that the government pays for research which a private organization can then benefit from without similarly paying back for it. However, it'd be even more silly to get worked up over every minor injustice when there are so much greater ones in equally easy reach.
Not a whole lot of those doing groundbreaking work in the field. And there aren't many interested laymen. On the whole, it seems unlikely to cause any measurable harm to the future if said laymen cannot read academic papers for free.
Now contrast the simplicity of interacting with the most complex web application and the comparative overwhelming depth and difficulty of making novel findings in computer science.
People use quite simple tools for complex interactions with other human beings (e.g. facebook). Someone interested in research may use quite simple tools like evince to view the latest research papers.
This is a persistent issue in all academia. I currently have access through a university, but trying to do research without someone paying your way is absurdly difficult. Of course the Internet is all about open-access, and I'm all for open-source software. It continually surprises me that academics don't worry more about this.