You think it’s not useful to know that there is a high degree of uncertainty and what the likely ranges might be?
I think it is useful. For contingency planning and the like.
Do you think that the estimate of 500,000 deaths in a "do nothing" scenario has been disproved? If so, how?
It looks to me that it would probably have been right within an order of magnitude, say 100k-1m. That's more useful than thinking it might be like a normal flu or wipe out 50%of the population.
What would you like to replace this sort of modelling with? Tea leaves, astrology perhaps?