I have never counted calories and never will, I go almost entirely by
feel.
I've been running an unofficial experiment with 20 odd years of trial and error, and what going by feel has led me to is a preference for fat and protein. That said, someone following the exact same diet as me would probably get fat. What I do works for me, but there's just too many variables for a once size fits all approach. A paleo diet is OK in theory, and is fairly similar to what I go for, but diets varied widely around the world, so "our" ancestors didn't all eat the same things, so different people are going to be suited to different things. Anytime you try to make everyone/everything conform to one golden rule, you're going to get exceptions that confound it, and that's the problem with nutritional science.
I'm convinced people must instinctively know what to eat and what not to - even dogs do - but they're overthinking it, and getting nowhere. Just look at all the nice people who claim "chocolate is good for you" - they can't possibly believe it, but they seem to think their own bodies will react differently if they know of a study that could conceivably support such a notion in a backhanded way.
Coincidentally the latest entry on a blog I follow deals with bacon - according to the "conventional wisdom" one of the worst possible foods one could eat, but one that I can't get enough of.
http://chaosandpain.blogspot.com.au/2014/12/bacon-superfood-or-hipster-bullshit....