Not trolling at all, and I certainly wasn't questioning the validity of the study. On the contrary, it sounds like a very thorough and well-researched piece of work with many valuable conclusions.
I was questioning the abuse of survey data in advertising campaigns. I think it's very misleading to use results from a survey like this to make claims like "1.2 million people have undiagnosed diabetes". The survey doesn't make that claim itself (as far as I can tell). It's an extrapolation from survey data that isn't warranted by the research methodology.
I haven't looked into the studies, but I'm willing to bet that the "undiagnosed diabetes" figure is based on the numbers of people who seek medical care for some other reason and are then found to have diabetes. This is the sort of figure that is readily accessible to NHS management and epidemiologists.