"People" isn't one person. Maybe most people choose sweeter yogurt, but that doesn't mean there isn't a substantial minority of people who would prefer less sweet yogurt.
Who knows if these tests even accurately measure what people would actually buy? I remember that Coca-Cola did such tests too once...
The guy I spoke with was unequivocal: they don't think they can sell enough to make it worth any shelf space at all. Their brand in my market gets allocated enough space for about five flavors. They know exactly what sells how much and where, and they know it to some extent for their competitors too (none of whom produce reduced-sugar, but most of whom do produce reduced-fat).
Now, if your argument is that these full time professional market researchers can't figure out what people would actually buy...well, I thought the guy sounded extremely intelligent and well trained, and he had no reason to deceive me.
Who knows if these tests even accurately measure what people would actually buy? I remember that Coca-Cola did such tests too once...