Okay, after reading Steve Rubel's explanation, I think I get it — it's just that there's nothing much to get. Or rather, there's exactly what I thought there was: OPML sharing will be used to determine what people are interested in. Consumers can find new sources of news and marketers can pitch to them more effectively.

This seems like a much bigger win for the marketers than the users, though. RSS empowered content consumers in some obvious ways — it made it much easier to stay on top of many sources of news. OPML provides a slick, hi-tech framework for providing statistically justified blog recommendations, but to be honest I've always felt that blogrolls and interblog links work fine for that sort of thing. OPML's big advantage will be centralizing that data so that it can be more easily mined.

I suppose it may help inject some more ad revenue into the blogosphere, but that's really the only major benefit I can see. I can't imagine a scenario under which an end user would be all that interested in using OPML sharing. This is a technology that'll be implemented in the backend of RSS readers in order to provide a shiny but unnecessary recommendation engine — and, not coincidentally, a new source of marketing data. But that seems to be about it.