I think that the best products out there are ones where the designer identifies a problem and comes up with a better solution. From here, sprouts a product that provides innovation. In the case of the "meeting of minds", I believe we are all seeing that the need is to produce and deliver effective training for little money. Within that need, there are many effective ways of doing it. I think that one really good way of solving the problem is to use web-based techniques, starting with Web 1.0 techniques. Of course, Web 1.0 requires a "design pattern" to be applied to training/education, but the approach should use the "best of breed" of both instruction and of web technologies.
I tend to rant about "Solution looking for a Problem" primarily when I see design rationale that was applied as follows.
A. I know how to do X (e.g. build a database)
B. I can deliver training from X (e.g. a database)
C. I can sell X for lots more money than other solutions out there (e.g. a simple web server)
Therefore, a solution based on X is the best solution to training needs. If you're going to use a database, it should be because of your storage/retrieval/reporting needs.
Unfortunately, from my experience, the database delivery systems (aka LMSs) are also the strongest voices in the creation of the specification such as AICC and SCORM. Perversely, before SCORM 2004, the course was the component that had to do all of the navigation, calculation of completion, etc. whereas the Database was just a "memory" for the browser. Restated, the multi-million dollar server/database system just stored results from the course, while the FREE browser the course was being served in had to do all the calculations. Even though the course author could specify mastery score to the LMS, I never saw an LMS that tied the mastery score to the student's score to come up with a "Completion" report. For that matter, I have rarely seen a decent report from an LMS that provides a breakdown of how each question was answered by each student, and whether there was a problem with a specific test question.
I feel the same way about PodCasts and Wikis. They are good tools (Podcasts for disseminating information, Wikis/Content Management Systems for collaborative documentation), but calling them pure training is definitely a stretch ("Solution looking for a Problem"). That's what annoys me so much about the "big" analysts pushing these as the future of elearning. I see most of these technologies as good resources, not training.