No, it's pedagogically un-sound and there's a problem in your line of reasoning; I'll let you figure it out.

I think its a rather clever idea. Its just a "your mileage may vary" kind of idea.

Sure, it might cause other bad habits, but at least they'll be better than the current bad habits of playing totally wrong moves.

The main problem with "play moves you can name" is that the people who would benefit most from it are least likely to know all of the named jumps/approaches. And there are some borderline cases - for example, does the 3-space jump from the third line to the fourth line have a name?

But in general Mef's suggestion seems quite sound. What's the point of learning the names of all these jumps, approches, tesuji, connections, etc. if not to highlight that there is usually something worth looking at there? We are perfectly comfortable telling beginners things like "broken shape is bad," "the second line is the line of defeat," "corners-sides-center", but it is far easier to find broken shapes, second-line stones, and center-facing moves in joseki than it is to find stones in a two-three jump relationship.

More generally, any advice that you give beginners, if it is more specific than "play lots of games and do lots of tsumego", is going to take the form of a general rule that overrides their weak beginner instincts. If they could already apply the rule without having some sort of faith in the rule, they wouldn't need the rule. For example, "give your stones five liberties" is a rule that is helpful for beginners who can't see that once all their stones have three liberties, they are exposed to painful forcing sequences, double atari, squeezes, and so on. "Give your stones five liberties if and only if giving your stones five liberties is the best play" is not a rule that is helpful in this situation; the problem is precisely that the beginner can't see why getting five liberties might be the best move. So he starts out following the rule against his better judgment. Then he grows comfortable with the rule, then he starts to get a feel for when his stones are safe because they have enough liberties, and eventually he can read well enough that the rule is unnecessary.

So any advice that requires you to go against your better judgment and try out something new carries within it the seed of a bad habit, if you prefer to rely on your comfy old rule past the point where you need it as a heuristic.

If you want to be pedantic, it's a daidaigema, but in the context of my suggestion "light extension" is fine, because if you know it's a light extension then presumably you know why you are making it (to extend) and have a plan for dealing with it if it is split (treat it lightly). Personally I think the fact the problem you list is actually one of the strengths of the advice -- as you get stronger and learn more named techniques, you have more tools in your arsenal. The idea is mainly to stop the "I don't know why I made this move" or "I don't know what this move is doing" type of plays.