Oodweyne,
You've written a lot but you did not say much. Instead of writing gibberish, why not give explicit example from the report? I can already tell by your answers that you do not have any background in methodology or statistics. The fact that you fixate on sample size instead of the statistical power, tells me of your methodological ignorance. Only someone with who lacks research knowledge will think that a bigger sample size will increase the power!
Anyways, since you're so fixated on the sample size: here they are ( I bet you didn't read the report).
more than 4 thousand household samples, and more than 21,000 individual sample size seems more than enough for me. That sample size is used to project to a population/household size which covers more than 80% of most regions.
The differences in sample sizes can easily be adjusted with a simple variance analysis. Your knowledge in statistics really dissapointed me oodweyne
Lastly, the report was prepared by Utz Pape, who holds a Bsc, Msc, and Phd in statistics. But I guess the words of Oodweyne should be heavier