In our LMS (Blackboard 8), instructors have the following drop-down options in a content area (in the order they appear): Learning Unit, Survey, Assignment, Discussion Board, Group, Tool, Document Package, Syllabus, Offline Content, Podcast Episode, Google Scholar Search, Google Scholar Content, Wiki, Blog.
All these choices are good, right? Instructors can use the drop-down menus to easily insert a range of different content and Web 2.0 items, making their courses richer while finding the right tool to meet their teaching goals.
The problem lies in the paradox of choice. Turns out that having too many choices often limits our ability to choose anything. A range of options channels us into the default. The Heath brothers, in their new book Switch  (which I'm reading now and enjoying), call this "decision paralysis." They tell the story of an experiment in which the likelihood that physicians would stick to a default plan (a previously decided up on surgery) increased from 28 to 47 percent when the possible alternative drug treatments went from 1 to 2. As the Heath brothers wrote in a Fast Company article  on the same topic:
"What happened here is decision paralysis. More options, even good ones, can freeze us, leading us to stick with the "default" plan, which in this case was slicing open someone's hip. This clearly is not rational behavior, but it is human behavior. Similar tests with different groups have revealed consistent results."
In the above mentioned LMS example, what we really want our instructors to think about is utilizing the tools that encourage collaboration, active learning and engagement. In my experience, the 3 most important tools in the LMS to meet these goals are the Discussion Board, Wiki, and Blog. Courses that use these tools to promote engagement, and use them properly, are almost always superior to courses that do not.
Yet, the design of our LMS discourages the use of these interactive tools. By giving so many choices the instructor is less likely to experiment with any one of them. By having so many options the instructor is less likely to try the interactive tools that make the biggest difference for learning.
The LMS seems particularly vulnerable to a kitchen sink approach to features. Someone thinks, "this tool is great, this feature is wonderful", and more and more options are added. What is lost, however, is a decision architecture that encourages good course design (and hence learning).
My advice is that before we think about implementing new features in our LMS systems that we should do some pruning. What are the features that are most important and what can go? How can we simplify our LMS tools so that the most important features are prioritized?
The paradox is that we've succeeded in increasing the options for teaching with our LMS while diminishing the probability that the tool will be used effectively.