I was interested in the study mentioned, as I'd never before heard of positive evidence for such a phenomenon. This [0] appears to be the paper referred to. The results seem pretty bizarre and hardly support the rosy view presented in the article. In particular, the symptoms are different for nearly every trial and symptoms were reported for half of the sham trials.

This is also 20 trials with a single subject, so it's probably too early to talk about broad applicability. It would also be good to see some replications in a more respected journal.

[0] http://www.avaate.org/IMG/pdf/EHS-_A_Novel_Diagnosis.pdf