Usability recruiting and internet accessibility?

Does not compute himage This morning, while recruiting participants for an upcoming startup usability study, I stumbled upon a mythical creature I didn't think existed. After chasing it through the underbrush of the internet to no avail, I eventually gave up, disgusted and disheartened, for the creature was gone - potentially forever. And the creature in question? That rare breed of human who doesn't have access to a DSL connection. I know, I know, breathe in deeply - I didn't think they existed, either. But it turns out that this kind of person is out there, and we have to plan for them accordingly.  More and more I find that people respond to my online recruitment ads, often for a remote study that needs screensharing and the like, without having the necessary technological hardware to actually complete the study. Like that story where from the band of monkeys eventually one will write Shakespeare, sometimes potential participants make it into the pool when they really shouldn't. It would be easy to blame the participant - I mean, who doesn't have DSL access these days in the U.S., anyway? As it turns out, quite a few people don't, so in the end it really is our bad as user researchers, not the participants. We need to make sure we're designing studies - and recruitment procedures - that aren't exclusionary from a technological perspective, even when we're building a product or service that requires (relatively) advanced technology. Examples of non-traditional recruitment abound: Etan once stood in a BART station for hours to recruit BART riders, while Mick loitered in a Best Buy (with permission of course) hoping to snag potential wireless hub buyers. The main problem, however, is that these methods, while often very effective, are time consuming - and expensive. So how to recruit in an internet-ready world, when we're not all on the same technological page?  Any thoughts? Examples?