Extreme Gene Transfer: How Widespread?

A theme I've been developing (clumsily) in recent blogs is that in the real world, DNA is shared between organisms, particularly microorganisms, across species lines (maybe genus, family, and other boundaries as well), rather more frequently than most people are prepared to believe.

Note to self: How would one determine how much free DNA (extracellular, non-viral DNA) is present in a gram of topsoil? Or a milliliter of benthic mud?

Hypothesis: Promiscuous, freeform DNA-sharing is a default behavior of (nearly all) microorganisms. The cell wall is a specialized organelle that exists to rate-limit this process.

Why make such a hypothesis? Two reasons:

1. Because it explains speciation (in microorganisms, at least) better than point-mutation trial-and-error.

2. Because it explains certain novelties of nature that are hard to explain otherwise, such as the recent finding of an entire bacterial genome incorporated in the genome of Drosophila. See: Dunning-Hotopp, Clark, Oliveira, Foster, Fischer, Torres, Giebel, Kumar, Ishmael, Wang, Ingram, Nene, Shepard, Tomkins, Richards, Spiro, Ghedin, Slatko, Tettelin & Werren, "Widespread lateral gene transfer from intracellular bacteria to multicellular eukaryotes" in Science doi:10.1126/science.1142490.