It can be done, but you have to make it a deliberate project and put some time and thought into it, rather than just stumbling on a bunch of programmers in a cafe.
The story about the programmers in the cafe was not evidence for any claim. It was just an explanation of what set me thinking about the problem. It's not a pillar holding up anything, so attacking it proves nothing.
(Incidentally, a moment's thought would have made it clear that I have in fact "put some time and thought into it." I've spent 24 years as a professional programmer, during which I've observed thousands of programmers working for all sizes of companies. Plus I've seen first hand the transformation undergone by roughly 200 YC-funded founders so far.)
The second half is fallacy:
But rather than pointing out fallacies, a better way to refute Graham's evolutionary argument is by reductio ad absurdum. ... Our ancestors lived in a world that was shrouded in darkness half the day, therefore we would be happier without electric lights.
This supposed refutation can itself be disproven by reductio ad absurdum, because it could be applied to any explanation of our inclinations based on how we evolved.
(Bad example, incidentally. There do seem to be conflicts between electric lights and human biology: http://www.washingtonpost.com/wp-dyn/content/article/2008/02...)
Neither of these arguments even attempts to refute the central point of the essay. In case anyone wants to try it, the central point is that in an organization organized as a tree structure, structural forces tend to give each person freedom in inverse proportion to the size of the whole tree. That we work better in groups of 10 than 100 I feel is obvious enough not to need justification. The argument from evolution is just an attempt to explain why groups that size work.
Though on the whole my reaction is "I want my 20 minutes back," there was one encouraging thing about this experience. Even these dishonest DH5s are more civil than Atwood DH2ing about my choice of metaphors.