You could try and employ R to unzip the data. I once had to use it because the unzip node from KNIME was not able to handle certain strange dots and underlines and empty folders in a zip file.
Here you tell unzip which files (csv) you want and it can also handle blank folders on my Mac seemingly.
