So first off the ranks is the predictable issues; firstly it's expensive business and costs a lot to produce. Lots of time, lots of money to get your end result. Chances are the better your elearning is the more it cost to get produced (although some really expensive stuff can be equally bad). True for sure but hey, if you've got the budget you just shrug your shoulders and accept it as one of the factors you need to take on board.
Secondly there's the interactions. Now poor and cheap elearning tends to have low level interactions so it's clearly an impact there, but if you've got time and money it's possible to have highly interactive high-quality elearning so surely this is a moot point? Not quite, because the interactions on elearning modules are pre-prescribed rather than real-life interactions. If I speak to someone virtually or in person there's a real interaction and it's 'right now'. If I had to think of possible questions someone might ask it would be more like a FAQ guide than real interaction. My point is that an elearning module is highly unlikely to have artificial intelligence in it and is far more likely to have interactions that were written at the time of making it. If you follow some of my other ramblings you'll know that knowledge is not a thing and flows and changes with time - as would your interactions of course, unless they were written and published at one time - then they couldn't...
Then there is the other side of the time issue. Producing good elearning is rarely done in a day or a week (I'll stop there but you get the idea). The problem with this is that in many areas things change quickly and if your elearning takes months to produce it simply can't meet the needs if they change at all (and they often do). Let's also wrap up into that the time of key staff needed to provide input, potentially resources, busy schedules and time is no small issue... but it's still not the biggest issue that I think exists, but we're getting close.
For me the biggest problem with elearning modules is that they are a finished product. If you subscribe to #WOL or my own theory #LOL you'll know that actually the finished product never really exists. What I mean by that is that we are essentially working in perpetual beta and that learning is an action or a verb rather than a noun, so if knowledge and learning change, so must the information. If elearning modules were never 'published' and you carried on working on them, they could adapt with time, change as they needed to and have more interactions added as they changed. Sure you can make a version two and publish that, but as well as the time (and cost) issues you may have it's still a version rather than 'live' if you know what I mean.
Of course, the saving grace is that traditional elearning is not the answer on its own anyway. We still (and perhaps more than ever) need people to support and connect to achieve learning. Those live interactions can work really well, particularly when we connect students rather than just teacher to students. In this way the interactive MOOCs are far more effective for learning than even the most polished stand-alone elearning. I'm not saying MOOCs are any more the solution than elearning modules though, too many are heavy video based non-interactive content and assessments with some human interaction, but theoretically that MOOC style learning is far closer to achieving our learning aims.
As always I'm not 100% sure this is the final idea and I know to some extent I'll change my mind on at least some of this over time - so please comment on G+ or on this blog directly or via Twitter (@the_nthdegree) or... you'll find me :)