> Even if we assume unconfirmed cases 5 times that, we're still talking single-digit percentages of the population infected.
Wasn't you point just that we shouldn't assume? I know that there are models looking at a possibility for only 1% of cases being confirmed. That's been in Sweden and not NYC, but just throwing a 5 out there doesn't make a lot of sense.
As far as I know (at least that was the case ~2 years ago) Spotify would be profitable if they chose to, but reinvested as much as they could into a higher market penetration insetad.