RE: assumption vs. research results: It would be quite a challenge to accurately measure how much a largely cosmetic change in a course might affect student learning. Personally, I've always thought that the quality of the slides sends a message to the audience: low-quality slides can have an adverse affect on student morale, motivation, and attention spans, while well-designed slides could have a positive impact by keeping students more engaged and perhaps even improving long-term retention.
Proving my assumption quantitatively as opposed to anecdotally, however, would be a challenge. I suppose I could teach two sections: one with a set of "cohesive" slides and the other with a set of "haphazard" slides, and compare student performance. Even if I conducted that experiment, it's often difficult to get reliable data that way. (Any difference in scores could be due to other factors, such as what time of day each section was taught, or the raw talent of the students in each section.)
About my remarks on the "quality" of my slides, there are several practices I often employ regularly; for example:
- Replace bullets with pictorial cues
- Add occasional humorous cultural references in strategic places
- Add discussion questions intended to launch an in-class discussion
- Pepper the lecture with active learning activities
Based on student remarks (via in-class comments and end-of-course surveys), I get the feeling that my efforts are appreciated, and it's worth the time I invest. Last year, one student put his hand up in the middle of a lecture last year, and asked me, "How much time do you spend on your slides?" He told me outright that he appreciated the effort. Another student in that same section was engaged to an intern who was working for me on a committee. She told me, "My fiancé says he really enjoys your class; it's more interesting than most."
As with many things in learning, though, it's all about tradeoffs. While I feel my slides help facilitate a more lively and interactive presentation, I've also been told that my slides are great during lectures, but not such a good resource for studying. Moreover, I'm an adjunct professor, so it's not like I'm committing time that would be better spent on research.
RE: The question at hand is whether or not refactoring old lectures is worthwhile, as they stand they are sort of hap-hazard and glued together of slides from various lecturers: I think it would be worthwhile, but maybe you could refactor them gradually over time, so it wasn't such an enormous task. Try to get them into a cohesive theme and smooth the transitions as you continually improve the course.
As much as I admire your thirst for hard data, there are times to go with your gut. If you cringe at the haphazard feel of your presentation, perhaps your students are noticing, too.
One last hint to consider: If getting the slides more cohesive is a daunting task, perhaps you can hire a student to do some of the heavy lifting. We did that for a course that needed a lot of material moved from written notes into PowerPoint. It was an efficient way to get much of the work done at a very reasonable cost.