Given the ratio of one black hole per 1000 stars, which is the best we can guess right now, the average distance to the closest black hole is 36 lightyears.
I don't know if we'd have expected to have seen lensing effects from our whole-sky surveys, but it seems to me that this would also be probable.
Finally, if you send a line through our galaxy, you'd expect it to pass near a lot of these black holes -- and thus you'd expect to see a lot of microlensing effects which the article states don't seem to occur nearly often enough.
Normal black holes are really small, much smaller than a planet, which we can barely detect now. And those planets must be orbiting a star and in the correct inclination to be detected.
Especially for the graph taken from http://arxiv.org/abs/1112.1320
