I think your question on evaluation is a very valid one. I've done a few evaluations but I've found, almost without exception, that the people whom one is never required to ask or consult are the people who receive the programs, which to me is glaring.
It's very difficult to do research within native communities. The risk is very high in this area. Passing through any sort of ethics protocol is lengthy and consent from the chief and council or the health authority is very problematic.
I agree that more work has to be done in that area because I have found evaluation is a really weak area, and as you say, things are probably done over and over again without it being really firm and clear that this is indeed helping people.
When we are asked for reports on whatever we deliver, we're not even sure they're read, let alone how far the recommendations we give might go.