Cheers Dave. I'll get back to you on the decentralised production thing - it's something I'm very interested in; see e.g. this post on adaptive landscapes (especially the second and third examples, Andean potato growing and Balinese rice management). I think it's a very valid question: can 'top down' plant science and distributed systems work together? Even if they couldn't, of course, I'm not sure that justifies trashing experiments, but that's a separate argument.

On the analogy: No, the analogy isn't perfect. They never are, though, are they? Though as I've said here, I actually think Andean potato farmers (and probably maize and tomato producers, also in central/south Americas) could probably claim to be the firs open source producers.

Do you mind if I repost your comment to the final blog version of the letter?