"Teaching" invented strategies?

I continue to think of an experience I had this year while subbing in Grade 4, the "new" math (resource that is) includes several invented strategies for addition and subtraction. Strategies that are introduced by the teacher, and practiced by the students, in a fashion that is similar to how we usually teach the standard algorithms. So then, I wonder, is this really an effective method of encouraring students to use invented strategies, strategies that we are encouraging students to use because they make sense to them, because they arrived at them through their own discovery? The textbook exercises go so far as to instruct students as to when they should use each invented strategy, and provide step by step rules to follow.......is this any different from a standard algorithm? I had the fortune of being present in the grade 4 classes for much of this unit, subbing for the different teachers at that grade level for several days each during the unit, I didn't however have the fortune of doing the planning, and so I went through the textbook exercises each day as assigned, and each day I was met with numerous blank stares, ESPECIALLY when it came to the use of the so-called invented strategies, the students were confused, and could not understand why any of these methods would be of use to them when they had already been using the standard algorithm (with questionable accuracy) for years. A few students in the class saw the efficiency and effectiveness that the invented strategies provided and were able to invent their own strategies by refining those in the book to make them even more efficient, deleting some of the "steps" and adding others that they saw as necessary, these students were making sense of the mathematics behind the strategies being presented to them, many of the others went through the steps outlined in the text, hoping that with practice they would somehow "get it". When the time came for the test on this unit the majority of students used the standard algorithms to solve problems where no particular method was suggested or required, even when it was reinforced that they could use whatever method they wanted, most used the  standard algorithm, though it was not always used correctly, and many of the errors made by students were similar.
This causes me to wonder, can we really "teach" invented strategies? I think not, I think that these invented strategies are best introduced to students by students, as they occur naturally. I think that practicing them is ineffective, as those who don't understand are unlikely to understand through repeated practice, but instead they require problems given in real life context that encourage and support the use of such strategies. They need to see that much of the mathematics we encounter in life is not pencil and paper.

Melanie