It's true that many kids, at a very young age, are curious about how things work. I see the biggest problem is keeping that interest alive, while teaching the tools they will need to do wonderful new stuff in the future.
That's a challenge. Among other reasons why, many educators at grade school and high school levels themselves don't know the relevance.
A simple example is the supposed "transition" from arithmetic, to algebra, to calculus. I know a whole lot of people, certainly the vast majority of students, who find these to be traumatic transitions. When in fact, they should be a natural process of generalizing what you had learned before. Those rules you were taught before now become special cases of something more general. Those equations you had to memorize now can be derived, and expanded, from basic and logical principles.
I know I wasn't taught this explicitly, and I'm sure most kids still aren't. I had to figure it out on my own. Kids who see these "transitions" as nuisances in their lives end up dropping out of STEM subjects.
In short, I'd say the problem is not that kids aren't exposed to science and engineering early enough. It's that they are turned off to it, when being taught the tools they need to pursue these interests.
It's true and more and more companies realize that the lack of engineers can bring them harm in future. During this year's CeBIT in Hannover I talked to Samsung people and they told me that in Korea they have special programming lessons even in Kindergarten. But they realize that at least for now it's impossible in eg Europe, that's why eg in Poland they started at primary school with children at the age of 6-7. Here's the site for it - you can use google translate to get translation from Polish :) http://mistrzowiekodowania.pl/
It's funny you should mention the transitions between algebra and calculus: when I was in high school algebra, we actually derived calculus. It wasn't part of the standard curriculum at the time, and I doubt I still have the mimeographed pages (which tells you how long ago it was).
I wouldn't be able to do it again at this point but I always remember it because that was what got me through my first semester of calculus in college. Without that background, I probably would have failed right then and there.
I posted a blog on EDN about a high-shool physics assignment where the circuit was impossible to solve. 30 comments so far. Remember, if you're logged in here, you're logged in there. Click on the image to see the full article.
Is not exposing children to engineering bordering on child abuse? I think you should get a parents permission first to be safe. That way if they become engineers later in life, you won't be held responsible. ;-)