The author’s argument is that the new, computerized evaluation forms are accurate. So whether the previous, paper evaluation was accurate is irrelevant.

The author’s conclusion is that the evaluations will accurately reflect what students think, not that what students think is accurate. Whether the students’ assessments are wise or foolish is irrelevant to the author’s argument here.

The author’s conclusion is simply that the new system of evaluation is accurate. Assuming that no university anywhere should use the old system would be far beyond the scope of his argument.

The author’s conclusion is an absolute one about the computer evaluations (specifically, that they’re accurate). So he doesn’t need to make any assumptions about the paper evaluations.

The negation is: dissatisfied students are more likely to submit computerized evaluations than satisfied students. If so, the evaluations would represent a biased sample, and might not be an accurate reflection of student opinion. This would undermine the author’s argument.