Obesity, skiny, tall, short, unibrow, hairy, flat figure

[STM1993]

Okay, fair point about the bullying thing.

I could agree with everything you've said, except here:

but its not unfair to use a statistical trend as a frame of reference for a group of people.

"No amount of experimentation can ever prove me right; a single experiment can prove me wrong." -Albert Einstein.

It's straight up wrong to generalize as long as there is a single exception; 5% means 1 in every 20 overweight people you meet everyday do suffer a genetic problem associated with it. None of those who were arguing mentioned anything about these few exceptional cases, but sticked to what was more like "Obese people have no excuse, and therefore they're all fat cuz lazy". I am not promoting genes as an excuse; all I am try to do is clarify that it's not fair to jump to conclusions every time they encounter a fat person.  

I think you're being too black & white about the idea of generalization, perhaps even fear the idea of it because of association with unreasonable extremes like racism. This isn't science where we're trying to prove a universal truth.

Let me put it this way: People in Germany speak German.
Obviously, not everyone in Germany speaks German - there are immigrants, tourists, the mute who can't speak etc, but would I be wrong to assume that the majority of people I meet there would speak German?
In the same way, if I'm looking at a large group of obese people, would I be wrong to say that the majority of them are obese because of bad lifestyle and not genetics etc when the statistics back this assumption?

Again, I want to stress that looking at people as a group and looking at individuals is a very different matter, largely because individuals don't have a sample size (plus the obvious emotional factors). That, and making a generalization does not mean declaring a universal truth.

EDIT:
Strictly speaking, I think positive/neutral/negative generalizations are equally valid from a purely logical point of view. Personally I find that the only reason why a neutral generalization is accepted while a positive/negative generalization is shunned is really because of the perceived implications or fear of offending someone - I don't want to believe someone is fat because he is lazy, but it still stands that the majority are.