This is something I've wondered over for a while as I like to peruse the paleo foodie sites for ideas. It seems to me that our ancestors would not have even had the means let alone the patience to separate the fat from the coconut or avocado or nut ... or even render fat. Ghee especially confounds me in that it's still a source of fat from a non-paleo source even if the proteins are gone.
So this one makes me curious. Why not just boiled/poached eggs? Grilled/baked/poached meats, poultry, fish & seafood?
They were actually pretty adept at bone-grease rendering: http://www.alexandriaarchive.org/bonecommons/items/show/1058
More evidence (free full text): http://www.u.arizona.edu/~mstiner/pdf/Manne_etal2005.pdf
From that link -
The second family of bone processing techniques emerged during the Upper Paleolithic (though not in all regions) and is a good deal more complex. Heat- in-liquid rendering (sensu Binford 1978; Brink 1997; Delpech & Rigaud 1974; Lupo & Schmitt 1997; Wandsnider 1997), also called “bone grease rendering”, allows a wider range of nutrients to be extracted per carcass. The technique is far more labor-intensive than cold processing methods, however. The advantages of the method are greatest where key resources are limited, whether due to unpredictable supplies or prey, constriction of hunting territories, or both. For this extraction technique, large amounts of spongy bone tissues of vertebrae and softer limb ends (Figure 3) are fragmented or pulverized, and then boiled in water by such methods as adding heated stones to the mixture. The heated fatty components float to the top of the mixture and form discrete, relatively pure layers of fat that can be skimmed-off and stored. Heat-in-liquid techniques maximize both the protein and fat yields per carcass, well in excess of what is possible from cold extraction techniques alone (Binford 1978; Lupo & Schmitt 1997; Wandsnider 1997). The purified fatty components can be stored for months (Soffer 1989).
Not to justify daily butter stick habits or anything, just to establish precedent for some added fat.
N-thing the bone grease processing, also there are many foraging cultures that process nuts to produce fat to add to food, such as dika nuts. There are also several highly fatty grubs in Africa people would add to meals.
Either way, I guess I do get most of my fat from bone grease since I use the fat from stock to cook with, but paleo is not about specific foods. If it were, I'd have to squeeze some grubs over my potatoes and I'm just grossed out by that, though I know I should be more open minded. Marrow on potatoes is good though! Paleo is more about metabolism and autoimmune awareness than eating foods that paleo people actually ate, but I think most people would benefit from eating bone grease, red palm oil, marrow, and brain as added fats rather than coconut oil or lard, since comparatively these are much more nutrient dense, though coconut oil has specific fats that can be quite useful to the body, particularly for lactating women.
Our ancestors had a large selection of grass feed meat and they ate portions of the animal that many of us no longer do (such as the fatty brain).
You need the extra fat to make up for the loss of very high carbs in a typical SAD diet.
I second Eric's answer, but I want to address your comment about ghee. Nearly all of the carbohydrates and proteins are removed from butter to make ghee. Those are also the things that make dairy "questionably" paleo. IMO ghee is a fine paleo food.
I don't think they are, and if they are I don't consider them optimal. Added fats, are in a sense empty calories, they are the white rice of fats, very micro nutrient poor. They do have some benefits I suppose, but I rather get my fats from whole foods ie. eating the coconut and not just the oil, or olives over olive oil, egg yolks, fatty cuts of meat, etc.
Take it from someone who use to dump coconut oil over everything just for calories.
I think the "dairy is not paleo" think can be put to bed.
True, it was not consumed during the evolution of the majority of the human genome, but that's not relevant. Signficant portions of the worlds population have evolved genotypes to benefit from the consumption of dairy.
Everyone have a listen to LaLonde's talk (for the theoretical perspective). Kresser has a great article as well: http://www.google.com/search?sourceid=chrome&ie=UTF-8&q=dairy+chriskresser