What does Sue Denim (reviewer of Ferguson’s code) mean is actually to defund the branch which stimulates epidemic patterns based on Susceptible-Infectious-Recovered-Dead (SIDR) model. It is sensible not because Ferguson's code is horrible or buggy messy, but because its outcome always has failed to match reality:

In 2002, it predicted up to 50,000 people would die from BSE. 177 died
In 2005, it estimated bird would kill 200m people. It killed 282
In 2009, it estimated swine flu would kill 60,000. It killed 457

Other researchers work on Coronavirus modeling (based on SIDR) also report absurdly high mortality. Of course, those researching are heavily influenced by Prof. Ferguson. The fundamental dilemma, in my humble opinion, is that mathematical estimation approaches based on SIDR model lack the insight into the nature of human immunity. Almost all basic parameters are poorly defined. The model could be complex enough. But eventually, during the pandemmic, we realize that it could notoriously naive.