Remember, it's a ratio. What matters isn't the raw number of missed cases, it's the ratio of missed-to-detected cases versus uncounted-to-counted deaths.
If Italy has missed half of its deaths but also half of its total cases, the fatality ratio would remain the same. To bring its CFR down to the level of South Korea, you'd have to make the implausible assumption that Italy has caught nearly all of its deaths but missed more than 85% of its total cases.
Yes, I know it’s a ratio. That’s why I used the term “numerator” and “denominator”.
It’s not at all implausible that Italy has missed the vast majority of it’s cases, because they’re barely testing, and they’re not testing minor cases. It’s essentially guaranteed that they’re missing a huge number of cases.
A reasonable estimate is that they’re missing 10 cases for every one they actually detect. The error in the denominator is much larger than the error in the numerator.
If Italy has missed half of its deaths but also half of its total cases, the fatality ratio would remain the same. To bring its CFR down to the level of South Korea, you'd have to make the implausible assumption that Italy has caught nearly all of its deaths but missed more than 85% of its total cases.