by Jeff Sovern
Here. But their definition of "few" still seems high. "Only" 4.2% of the consumers whose homes were foreclosed upon are said to have been "harmed" by robosigning. That seems to come from a report from the OCC which has historically been captured by the banks, though since the appointment of Thomas Curry as its head, seems to be less so. And the articles notes that "some auditors" have found much higher error rates.
I suppose it all comes down to how you define “harm.”