In the start of the article they list assumptions. They do not assume that treatment can change their assumed 4 day asymptomatic state, nor their assumed 8 day symptomatic state. If Zn+Coax or Massive Vitamin C reduce either one of these, it can throw the model off and result in overprediction of cases. Alternatively, it could also be used to explain the decrease in transmission.

I also thought is was curious they chose 4 days and 8 days for the time period when a person is contagious. It seems that earlier I heard that 14 days was the asymptomatic period in which people were contagious. That would really change things from earlier models as well.