"it clearly has loads of potential for one key reason: it actually feels fun."
This is an excellent example of Programmed Instruction. Most of us probably learned typing the same way, and maybe even Logo. The reason this approach is good is because, in addition to being fun, it's also compatible with the Keller Plan:
>Granted you can get only so far with this approach, but it's still terribly underused in the current scheme of things.
In the teaching of concrete skills, I am not sure what the bounds are -- the burden on the instructional designer to administer the skill dependency graph as the skills become more complicated continues to grow, but I don't see that as a limiting function.
The problem is that concrete skills are only a small part of learning. This sort of approach is great for teaching kids to pass math tests, but less great at turning kids into mathematicians.
Ah, I'm glad you agree. I suppose my reaction is that most people underestimate just how far you can go with this approach, and I am glad to see that you are not one of them; we agree that you can train someone to solve calculus problems, but not the skills of derivation.
Where we seem to disagree is about wether this approach should be used in isolation. Frankly, I think most people (successfully) trained in theory are bored by application so there is a use in multiple educational paths.
While application is a small subset of the available learning to someone interested in a field, it is often a sufficient subset to be a productively contributing member of society.
This is an excellent example of Programmed Instruction. Most of us probably learned typing the same way, and maybe even Logo. The reason this approach is good is because, in addition to being fun, it's also compatible with the Keller Plan:
http://en.wikipedia.org/wiki/Keller_Plan
Granted you can get only so far with this approach, but it's still terribly underused in the current scheme of things.