Assessing our instructors, or ourselves as instructors is a messy business. Measuring student output (grades) can be interpreted too widely. Is an instructor with a high classroom average trying to buy positive reviews, too tired to give constructive feedback, or just a really effective teacher?
There are a few things going on here: are the student surveys effective? What are they measuring? Are students taught to provide effective feedback ever? If something works one semester, resulting in positive evaluations, why can it fail in a following semester? If there are that many variables, is it worth wondering if paper or online surveys are better, richer, or more timely?