This study needs to be interpreted with an abundance of caution, given we already have real-world evidence that contradicts this. Results heavily dependent on assumptions- & these don't account for important factors e.g. poor compliance with isolation. https://www.bbc.co.uk/news/health-55980052

We know that during the last lockdown, when schools were open the variant was spreading with an R of 1.5, and even when the variant wasn't yet at the high level it is now, R was above 1 in regions where this was increasing in frequency, despite lockdown.

We shouldn't expect things to be different unless we make changes- unless we make an effort to mitigate spread in schools (interesting that the model doesn't really assess the impact of mitigatory measures - masks/ventilation/smaller class sizes, except in terms of rota learning)