>Take deaths, multiply by 400 and that's how many cases you have. This comes from the assumption of 1% mortality and 3 weeks to die.

So that 400 multiplier gives the number of cases as of 3 weeks prior to the death count used in the calculation?