Yes, it is best to observe the user in their natural environment using a product. That is not always possible. I worked on a pregnancy tester and all of the women did not want me to watch them pee on a stick. I worked on a glucose meter. I don’t want to follow these people 24/7 to see them test first thing in the morning and when they are out late clubbing. HIPPA restricts a lot of medical device ethnographic research. Then there is the economics. Getting 12 people to one spot costs less than going to 12 different places. That’s why docs don’t do house calls. As for one-on-ones versus group thinks, there are advantages. People can feed off of each other, kind of like a brainstorming session. In both cases, the success is usually determined by the facilitator.
So I will do my best Donald Rumsfeld. Should you do product research? Yes. Is observing the customer use the product in its intended setting the best methodology? Maybe. Are focus groups worthless? No.