Outcome of critical cases is heavily reliant upon whether the medical system can handle the number of infected. If someone requires a ventilator and you're out of ventilators they're kinda fucked. There's also evidence that lots of people show little to no symptoms. There's probably a large number of people that simply don't know they had it. Which is both good and bad.

If we manage to keep it from spreading too far too fast, from the variety of studies I've seen then anywhere from .5% to 1% is probably more realistic. But if we don't, expect something higher. This is why waiting for "enough" people to die as evidence of severity, then responding is a poor response.

FWIW the Spanish Flu infected 27% of the world's population. 60-70% seems really high.