Personally, I've always interpreted the word "appropriate" to be aimed at correcting other first-worlders with (often dangerous) misconceptions about how development works. Case in point: your water system example.
Let's say you're an organization interested in improving the water situation in a rural village in Kenya. Which of these two options makes more sense?
1. A well that the village residents can build and maintain themselves with relatively small amounts of initial and ongoing costs, and that doesn't require specialized knowledge to keep running...
2. OR, a modern chlorinated water system, which depends on tons of pipes, valves, pumps, etc., and will require a massive construction effort to build, and will also need a steady supply of chlorine, and somebody who can fix all the pumps when they break, and somebody who knows enough about water treatment to keep the the water system from accidentally giving everybody cholera, and a steady supply of spare parts, etc. etc. etc.
I mean, yeah, in the best case, a modern chlorinated water system is a great thing... but you have to look at the whole picture and at the long term. A simple solution like a good well can have a real impact on a lot of people's lives, and unlike a proper water system has fewer expensive and dangerous failure modes. Oh, and by the way, the well solution was cheap enough that you've got enough money left over to go and build a well in the neighboring village, as well. Woohoo!
I agree that the term "appropriate technology" can be used in the matter in which you describe; however, to me it seems pretty obvious to say that technology is not a one-size-fits-all thing. There's no point in (for example) shipping a bunch of MRI machines or surgical robots to a rural clinic somewhere when what the clinic really needs is more rubber gloves, syringes, and malaria medication. In my mind, that's what the phrase "appropriate technology" is supposed to mean. Here's Steve's three-step process to a successful technological development intervention:
1. Actually learn about the people you're trying to help, preferably by living with them for a period of time. Learn about what resources (physical, educational, human, etc.) they have access to, and what resources they don't have. Learn about the problems they're facing in their day-to-day lives, and find out if the things you thought were problems going in actually are perceived as such by the people living with them.
2. Together with your new neighbors and friends, decide on a problem to try and address with a technical solution.
3. Decide on a technological solution that takes into account all of the stuff you learned in step 1 about what resources your population does and does not have access to.
The biggest and most common mistake that people working in development can make is to blindly throw technology at a problem, pat themselves on the back, and go home feeling good.