I'm wondering how many people think that rapid content development (RoboDemo, Breeze etc.) is yet another wrong direction for elearning? If high quality, well designed elearning is predominantly under-used, with very patchy evidence that people actually change as a result, aren't we just likely to end up producing the same useless stuff, but just much quicker and cheaper? But cheap stuff that doesn't work still doesn't work.
Actually - I'm in two minds about this; perhaps by widely diseminating the tools to produce elearning we can develop a critical mass of understanding in people who are currently described as "non-specialists". Maybe embedding learning in organisations can only really occur if it's produced, as well as consumed, at the point of need.
On the other hand, we might just confirm the hugely widespread misunderstanding of quite how hard it is to really get people to learn things - thereby losing out in two ways 1) our expertise is undermined 2) our reputation is sullied.