In Sweden it's a left-wing trope (the right wants lockdown and masks while the left supports the lax approach so the model has mostly received criticism from the left) and I believed it but you're completely right, so I apologize:

https://twitter.com/imperialcollege/status/13077606155456061...

> Thanks for confirming: Neil Ferguson’s Imperial model did not point to 85,000 Covid deaths for Sweden; another researcher did. A mean time-varying reproduction number Rt estimate in March is not a prediction of deaths nor is it a model for predicting them.

Ah, I should have been more specific that I only follow this in the English-language media so I’ve seen it newspapers in the UK or Australia (the US discourse doesn’t mention this much).