Although I am friends with David (the author of this blog post), I do not agree that open access is fundamentally the wrong model, but his argument is not without merit.
I'd guess that a "teaching university" would have nearly the same need for access (if the teaching is any good), but much less need for submission, compared to a "research university", so the teaching university should benefit from shifting cost from access to submission.
Publishing in ACM and IEEE conferences are essentially author-pay. Technically, you don't pay to publish in the conference, but you must pay to register for the conference. And most conferences have a rule that if you don't register, you can't publish.
There are many objections to the ACM and IEEE practices about maintaining paywalls, but I think that as professional societies, their policies can change. The for-profit journals, on the other hand, probably cannot survive if they allowed such open access. And nor should they: their business model is no longer valid.
* The review process for conferences is necessarily compressed relative to journals. This usually lowers the quality of reviews and also limits paper length.
* It is difficult to "grow" a highly selective conference, so the acceptance rate is usually quite high. This means that archaic tenure review committees evaluating on where you publish instead of what you publish will ascribe little weight to your conference papers. If people are not submitting their best work to a conference, it can't afford to be overly selective. This is much the same problem that new journals have, but worse because conferences usually have to reach a critical number of people to be considered a success.
This means that archaic tenure review committees evaluating on where you publish instead of what you publish will ascribe little weight to your conference papers.
Not in CS, in my experience. People in CS know which are the highly selective (what I have always heard called "top tier") conferences. In my experience in industry and academia, conferences are our primary consideration and journals are an afterthought. (Sometimes not even a thought.) The CRA memo is from 1999. I think most people have caught on by now.