// Fiction, I was allowing for another 100,000 deaths, which probably won't happen. //
Why not? At the moment the world is adding officially around 5,000 deaths per day, so by the end of the month alone, at the current rate, we should see more than half a million deaths. Plus, the official count is if anything an underestimate. Firstly, there are some countries who we can, sadly, safely assume are straight-up lying about their death count: North Korea, China, Iran and Russia being the most obvious.
Secondly, the way in which deaths are counted varies across nations. Russian statistics, in as much as they are honest, are based purely on autopsy results, rather than on deaths among those who tested positive and had Covid-19 mentioned on the death certificate. In the UK, the criteria for making the official death toll has changed at least twice: initially it only included deaths in hospitals among positive cases, then expanded to include deaths in care homes, and then earlier this month expanded to include deaths among those who tested positive where the test was carried out by a private company (as opposed to an NHS test). And even *that* is likely to be an undercount, because the the ONS has recorded at least another 10,000 deaths in all settings where Covid-19 was mentioned on the death certificate (1). This is not to criticise the UK's statistics, because it's the same in many other places. Spain has reported 27,000 deaths, but required a positive test in order to count the death, and so there's maybe as many as a further 16,000 deaths that should be added to the official toll (2).
Thirdly, it may be that the country simply doesn't have the testing capacity to track Covid-related deaths accurately, so undercounts by accident rather than design. This is likely to be true in Latin American countries.
All told, we probably breezed past 500,000 Covid-19 deaths already some time in early May, and when the final count comes in I wouldn't be surprised at all to see a death toll in the millions.
And, finally -- and again, I cannot stress this enough --
not everybody on the planet has had Covid-19 yet. It's meaningless to take the entire human population as your denominator in assessing risks. It's almost certainly true that more people have contracted the CoronaVirus than there are confirmed cases, but that isn't the same thing. In the UK it's estimated that somewhere around 10%-20% of the population has had the virus at one point or another (possibly an underestimate), which leaves around 5/6 of the population as yet untouched. The disease still hasn't really taken hold yet in Africa, where they haven't yet recorded more than 200 deaths in a day across the continent; it's only just starting to take off in India (3); the situation in Brazil is probably yet to peak (4). There are many more people yet to be infected, and many more people, sadly, who are going to die before this is over.
(1)
https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/datasets/weeklyprovisionalfiguresondeathsregisteredinenglandandwales
(2)
https://elpais.com/sociedad/2020/05/27/actualidad/1590570927_371193.html
(3)
https://www.worldometers.info/coronavirus/country/india/ -- notice the exponential growth curve in daily recorded deaths
(4)
https://www.thetimes.co.uk/article/coronavirus-deaths-soar-but-brazil-has-to-lift-lockdowns-says-president-bolsonaro-lrn67xhk2