Something happened with my previous reply, so I'm trying again, hope this doesn't result in a double posting:
The 30% nr comes from an as of yet unpublished study by Tom Britton, and is purely hypothetical (as is the case in all epidemiological modeling).
Se hur matematikprofessorn raknar ut stockholmarna nar uppnar flockimmunitet | SVT Nyheter
The initial antibody tests were performed on samples from 200 blood donors, 11% of which came back positive. Since blood donors make up a highly constrained test group, senior dr Jan Albert who oversees the whole thing has argued that the actual nr of infected is likely to be much higher.
Nytt test: Minst 11 av 100 har antikroppar mot coronaviruset i Stockholmsomradet | SVT Nyheter
Both these gentlemen appeared on the news yesterday and explained how they think their results are reasonable initial estimates and well aligned. So, the figures I gave are highly uncertain, which is why I used the words "suggests" and "preliminary" - maybe I should have been more up front about that. However, the rough IFR calculation is in line with the guesses we've seen from most (honest) experts in the field, such as Johan Giesecke, who gave a brilliant interview with UnHerd on YT a couple of days ago (I'm sure someone has posted that here already). I will get back to you with published results when testing ramps up if you're interested!
Santana: I'm not in the trenches at all, I'm currently writing a bachelor's thesis in Environmental Technology.