As an example, in Italy, 12 percent of all detected COVID-19 cases and 16 percent of all hospitalized patients were admitted to the intensive care unit.
I may be wrong, but I suspect the infection rate is much higher than the detection rate. To the extent this is true, percentages are misleading. However, I think there is a solution to this, which is to use absolute numbers. After all, the numbers of beds in a health care system is approximately known, so we can work with absolute numbers and make judgements of health care availability.
How many ICU beds with full respirator capability does Greece have? Those who become ill enough to require full respiratory support but don't get it simply die. Not all those who get full respiratory support survive. When there aren't enough places available those with the lowest chance of survival, i.e. people like me, will simply be left to die if we are unlucky enough to catch the virus and develop viral pneumonia/ARDS.
I could be wrong - I stopped following daily news back in mid-Feb, so I've missed a lot of this, but I've also missed a vast amount of hype - but I understand CV19 is dangerous to people who are over 70 and frail. People vulnerable enough that it doesn't take much to be fatal.
Assuming for one moment this is correct, we can observe the number of people in this catagories in any country is in the "millions" range (tens of millions in the USA) and there's no health care system in the world which can provide support for any reasonable fraction of them becoming significantly ill concurently.
Your earlier observation about a pandema being absolutely fine until it's not I think correct, and of course it echos the description of bankruptcy : it goes very slowly, and then all of a sudden.
The numbers, while still uncertain, are known with a lot more certainty than they were a couple of months ago and they don't make pretty reading. The vast majority of countries seem to fit a curve of doubling every 3 days. That means if you start with 1 case the whole of Greece becomes infected in 70 days. Initially Greece seemed to be on a doubling every 2 days which means if you start with 1 case the whole of Greece becomes infected in 46 days but it does seem to be having some success in slowing the rate, i.e. increasing the doubling time. Obviously there must come a point when the rate slows down simply because virtually everybody you meet will already have it but that only happens when a very large proportion of the population is infected. Also Greece is low down on the curve at present so it is difficult to tell exactly which curve it is on. This graph illustrates the situation very simply.
My take on this is that, because there is no vaccine, it's similar in nature to the Ebola outbreaks in Africa.
When you have a vaccine, you give it to everyone, and that solves the problem.
When you do not have a vaccine, and this I understand is how the Ebola outbreaks were contained, you have to identify everyone who is infected and isolate them. You do this firstly by mass testing and secondly by tracing back the steps of everyone you find who is infected, and isolating everyone they've been in touch with.
If you do not do this, because there is no vaccine, everyone and their dog is infected, and the nature of the rate of infection is basically that of the "doubling problem", something all software engineers know without thinking about because they work with base 2 all the time.
Large scale self-isolation, closures of public places, etc, can support finding and isolating by reducing the rate at which contagion occurs. There's a contest between the rate at which you can test and isolate infected and find those they've been in touch with, and the rate of contagion.
If you do not attempt to find and isolate the infected, or if you *cannot*, because it's too late and the numbers of infected are now too large for the resources available for this approach, then you are accepting that all or almost all people will become infected, because there is no vaccine.
Where this is a "doubling problem", and here we're talking about the vulnerable groups being the old who are weak, we're talking millions of people and there will be a sharp transition between "okay" and "not okay", when we go in a couple of days from a number of infected which can be supported by the health care system to a number which cannot be supported by the health care system (and which a couple of days later becomes a number much, much larger than can be supported by the health care system).
So, logically, as far as I can tell, what need to happen is a massive effort now to reduce the rate of contagion (closure of public places, curfew, quarantine for newcomers, etc), and a massive test, detect, backtrack and isolate effort, and then once the rate of new infections is basically zero, lift the restrictions (except for quarantine for newcomers).
It's that or a significant fatality rate among the entire population of frail over 70s. This entails their deaths, massive and widespread emotional loss, and the resultant economic disruption caused by people dealing with this.
I think given the infection rates I've seen, and the lack of mass testing, most or all Western countries have (not particularly deliberately) opted for the significant fatality rate.
I have also noticed that many web sites that are promoting the idea that cures have already been found are typically quite strongly anti-vaccine sites. That worries me. Everybody has an agenda. Russia is now strongly promoting the idea that the USA and the UK deliberately developed the virus and infected people in China with it. There is a lot of fake news out there.
Best thing to do is get some facts (the tricky bit - a good bullshit detector is required), and then think for yourself.
One irony here, if you'll permit me dark humour, is that the dead in the UK will be disproportionalty Brexiters! if this had happened a few years ago, Brexit would not have happened!