Joseph this is a really good post.
PURPOSIVE sampling is what we need in a qualitative study. Statistical sampling, stratified or not, is NOT what we need. Purposive sampling means you want to get a really broad sample that will tell you the RANGE of attitudes/experiences. Some people will think X, some will think Y and some will think Z. After you've done your qualitative study on a purposive sample, you might THEN take a statistical sample, perhaps stratified by age and gender, to address questions like "What % of women aged 20-30 think X?"
Does this make sense?
