It is interesting, but I don't buy some of it. There are complex variables involved, and that does come out a bit in the article, but other things they say are questionable at best. For example, the concept going around about Italy having the second oldest population worldwide is a bit misleading. They are referring to % of the population that is over 65,
for which Italy IS second (after Japan). However third on that list Germany, actually has more people in terms of headcount that are over 65 than Italy, as do countries like USA, Indonesia, India, and about 10 others, which means that in terms of people over 65 being in a higher risk category, other countries like Germany have more prospective deaths in pure numbers. If you look at the page I linked to you'll also see that there is another way of measuring age, in terms of life expectancy, and using that measurement Italy isn't in the top 10.
Another point is the way the article goes on about how things are counted, and how reliable the tests are. Are there a bunch of asymptomatic carriers in Italy? Is the Italian death count higher because a death is counted as a COVID 19 death if a person dies with the virus, without effort to ascertain if it was BECAUSE of the virus? Either way these questions kind of duck the most important issues. In terms of how many people are confirmed to have the virus and whether Italy is under counting this figure, that only really effects any proposed lethality percentage number, for which we are probably never going to have a really firm number for anyway. On the second question, how can you really tell if someone with an existing heart condition would not have lived another 10-15 years if they hadn't contracted COVID 19?
What would be more relevant is to see how much hospital admissions have gone up in areas of Italy where COVID 19 has spread. I've not yet seen any data on that, but it would be much more interesting to see if the arrival of the new corunavirus in Italy corresponds with a big increase in hospital admissions. This information would get around the kinds of easy to misinterpret questions being raised in the article.