It's hard to know how to compare evaluations of teaching with those of research. Take the description of my blog, "Scholarly Writing and How to Get it Done." It wouldn't make sense to have a blog about "Class Preparation and How to Get it Done." You have to show up prepared for class, and the preparation is an inherent part of that. The students might feel a course is more or less effective than another one, but each course exists and was taught by its instructor.
I just turned in my annual review form for '10. I had three articles appearing in print, and one review. Two more were submitted and are in press, along with one book review. For last year's review (2009 calendar year), I had two books and some other publications. Another year I might have zero articles appearing, or only one. It's easy to see, therefore, why research almost always ends up being valued more. The difference between four courses in which I get 4.2 on a five point scale, or 3.8 on the same scale, is minimal, compared to the difference between publishing three articles and zero, or between two books and one book thousand-word book review.
Teaching, judged by a qualitative measure of student satisfaction, cannot possibly compete on an equal footing with research, judged by quantitative measures of production. (Especially since research is not judged by sheer quantity, but by quantity of work accepted and published in reputable places.) There is simply no symmetrical way of judging the two activities. The difference between an excellent researcher and a non-productive one is immense, whereas the difference between a great instructor and a pretty good one is relatively small.
So any institution that values research at all will put too much emphasis on it, simply by giving it a value at all, because differentials of effort and accomplishment will show up in research much more strongly.
To give teaching equal weight, by this logic, is to devalue research. The only way to give it equal weight it to count relatively small differences in student satisfaction more than huge differentials in research productivity. I might produce five or ten times more than a colleague, but that colleague will never teach 10 times better than I do.
Many people do not understand this. I saw someone commenting at the Chronicle of Higher Education that professors need to take courses in pedagogy. I don't know anyone in my university who cannot teach a course. I've observed many people teaching and taken many classes when I was a student. On the other hand, some people do have difficulty making a contribution to a field of knowledge. People who can do that successfully are those who deserve to be in jobs that expect research.