Found an interesting article (through slashdot) on how we should be educating engineering students. The bird's eye view is that students should come out of school with the ability to continue learning rather than some specific skillset. The slashdot discussion and the article really cover mostly different ground on the subject, with the comments on /. debating the pros and cons of teaching engineers as thinkers or trades-people and the original article focusing more on how the education of thinkers can be done. Of course this isn't limited to engineering, any really good Comp. Sci. program runs the same give-and-take between teaching students to program and teaching them how to solve problems.
I don't really think there's necessarily a right answer in any of this. At this point in time it seems to me like the market for Engineers as workers with a certain skill set is notably present (and I'm going to entirely avoid the debate about whether that is right or wrong), while there is undeniable value in being able to apply knowledge outside of a specific skill set (I'm also going to avoid debate about whether this can be taught).
My own career, thus far, has been a case study in why problem solving is more important than specific skills. I am a graduate of the Computer Engineering Co-op program at the University of Alberta. Of my five terms of "Work Experience" I spent the first three doing PLC programming, circuit diagrams (in AutoCAD), and specification and ordering of parts for industrial control systems. My last two were spent doing data acquisition programming in Visual Basic in a direct precursor to my current position, which is a mix of Windows programming, putting together custom experiments, and heat transfer research. I'm also preparing to begin a M.Sc. degree in Mechanical Engineering.
The common theme in all of this is that none of my day-to-day work heavily involves anything I learned as a skill in my undergraduate degree, though I admit I haven't followed an entirely traditional career progression. I heard a Faculty member whom I have a tremendous amount of respect for answer the question "What qualities do you look for when you're looking for graduate students?" with something like "I like to see how they react to problems outside of their comfort zone." Of course he elaborated more on that, and the point was that it much more important for the student to be able to figure out something than to recite something. He also noted that you can get very different solutions to a problem from someone who solves it using a well-trained skill and someone who solves it with ingenuity, reasoning, and research.
Getting back to the article, I really think the term "meta-skills" is fantastic. While you learn skill in a program (and that's inevitable, most concepts in applied science and engineering are taught through some usable skill) the important thing isn't necessarily the skill or even the concept (which set the "trajectory" talked about in the article), but the ability to turn learning skills and concepts into a skill in itself. I think the article essentially skips the most important part of making that happen for students, telling them directly what they should be gaining. All the hand-on activity in small groups in the world isn't going to help them make that leap if they just see it as another lab assignment in a course that's only marginally (if that) related to the job they hope to have when they graduate.
As for the slashdot discussion, while interesting, I think it misses the point by focusing on theory vs. application. As many comments correctly point out, both theoretical understanding and practical ability are important, but in the context of the articles theories and concepts are simply another skill learned. A high level mathematics theory is every bit as narrow as a single programming language if the person who knows that skill doesn't have the ability to work outside their "comfort zone". Versatility is key. But like I said, the discussion is pretty interesting. I've noticed quite a few interesting comments. As a note on that last comment, it is fairly similar to (albeit longer) the engineering program here. All engineers here take a common first year, specialize in years 2-4 (while still taking cross-discipline courses), then have to practice as an E.I.T. for years (a residency of sorts) before being able apply to become a Professional Engineer (which is a term whose use is legally restricted).