hi doma,
i was thinking of a drug's rate of success in a population, not of the longevity of its use in an individual.
if information affects the efficacy of a drug and the value of information depreciates over time in a population, then according to your model i think that the efficacy of drugs should always decline in a virtual reality universe.