add another flux scaling strategy

As I explained, this adds the possibility to re-scale reactions bounds instead of using coupling. I also changed the default parameters for coupling because I think Federico used 0 instead of 0.01.