There's a correlation/causation issue here.

I do think telecommuting causes some increase in satisfaction, but I bet most of the higher satisfaction measured is the result of generally good management. Good management also causes telecommuting, resulting in a telecommuting/satisfaction correlation.