1. There is simply a cultural boundary between the geeks and the non-geek population that would cause such an idea to fail. I suppose there is a self-selection process involved such that geeks who are willing to help will have a greater natural ability to communicate to non-geeks than those who aren't. Then again, the non-geeks who are most likely to give Linux a try and call in may have higher-than-average technical skill and therefore the gap would not be so large.
2. If the gap were large (given that these are Linux geeks volunteering, above all), how would we incentivize geek retention? This would be a challenging, time-consuming job that demands a high amount of patience--much more than CD burning--for no reward besides karma and the satisfaction of a small handful of potentially new Linux users.
I am fan of Canonical and what they have accomplished with Ubuntu. They could spearhead such an idea, if they haven't already. Obviously, they would be inclined to support Ubuntu, but any Linux distro beyond that is probably too much for non-techies. A real financial backing could make a significant difference.
Not trying to bash your idea; like I said, I like it. Rather, trying to provide some criticism to evolve it. Anybody else?