I thought (from the Motherboard article, not the paper) that 1Hz was the resonance frequency of the vertical mode, but it was being forced at 0.6Hz, which would imply the initial vortex shedding was at 0.6Hz.
In the explanation of flutter for aircraft, it is said that it is necessary for two modes to become coupled. I'm proposing something similar is necessary to explain the bridge. It would still increase monotonically with wind speed, if it was true flutter, so would not be a resonance phenomenon.
My original comment was maybe unclear but it seems the torsion was forced to 0.2Hz, and 0.6Hz and 1Hz are both integer multiples of that. I proposed one or more of these became coupled. Possibility 1: coupling to 0.6Hz. Since that was presumably supplied externally, this could be called resonance. This seems unlikely though due to the evolving aerodynamic profile once torsion starts (i.e. there would no longer be a 0.6Hz input). Possibility 2: coupling to 1Hz. This sounds more like the aircraft version, and would increase with increasing wind speed.
I should probably read one of the papers.