Unless they are performing large scale sampling of the general population, the denominator (cases of confirmed influenza) is potentially much larger in practice.
Moreso because it seems likely that the confirmed cases are sampled from the sickest population seeking treatment.
And this article is further evidence of exactly this, since the "silent infections" here were previously uncounted.
We both understand that pandemics don't happen under controlled, fully observable situations, yes?
And therefore must be studied and predicted based on incomplete information?
And this has been the case since the dawn of epidemiology?
https://en.wikipedia.org/wiki/Influenza_pandemic
Being surprised by asymptomatic cases suggests their IFR numbers are high on the page you linked, due to low occurrence and conflation with other colds (or asymptomatic cases).
Is there anything to suggest that the asymptomatic strain detected here is worse?