What you're doing is cherry-picking the reasons why the mortality rate might be higher than reported, while ignoring the reasons why it might be lower.
Of course the number of deaths will go up - but so will the number of infections - both reported and actual. (its actually 21 deaths 'out of' 1140 in the UK now...) - and its very well established that the deaths are more or less confined to people with existing conditions. If you've seen up-to-date figures on what proportion of the reported infected are in the "at risk" groups, do tell - last I looked many were reported to be "self isolating at home" and are likely
not in the group at significant risk of death, who will be receiving hospital treatment. .
Couple of important quotes re. the UK situation:
"However, the government's chief scientific adviser Sir Patrick Vallance has said there are probably between 5,000 and 10,000 people infected in the UK at this time."
"He also said that many people will no longer be tested if they are showing mild symptoms. Only those with the most serious symptoms will be given a test in hospital."
Explore the data on how the coronavirus pandemic unfolded in the UK.
www.bbc.co.uk
There's no wishful thinking here: the "known cases" are
known to be far less than the actual number of infected and the testing program is
known to be skewed towards the more serious cases. Both of those will amp up the naively-calculated mortality rate.
Any "mortality rate" based on "reported cases vs. deaths" is hopelessly unreliable - as is the infection rate if the only people being tested are disproportionately the ones who have developed serious symptoms because they are particularly susceptible.
As I said - note of this is a justification for complacency, but over-reaction has a body count, too. E.g. more people will die if the health service deteriorates because half of the doctors and nurses are staying home to look after the kids because the schools and childcare facilities have been closed... because the politicians are under pressure to
do something.