Sweden as of 3 sept (latest data I can find) has a positive case rate of 11.1 per million of population, well below the government quarantine level for travel. However government probably won't take it off the list as to do so would be admitting its approach has worked. Sweden has consistently been below this level since mid august.
Swedens chef epidemiologist Anders Tegnell admitted that they didn't protect care homes well which contributed to the higher death rate, but then neither did we. He has admitted they could have done better to protect care homes.
Swedish deaths and ICU bed occupation now down to a record low (as is ours).
Immunity is increasingly being seen as not just a case of antibody levels, its more complex than that, T cells are a real big part but you cannot reliably or cheaply measure them in the same way as antibodies. Therefore just quoting antibody level studies is not definitive proof about population immunity. Check out the work of the Karolinska institute in Stockholm on T cell immunity - also there is increasing buzz about cross immunity from other coronavirus infections.
I wish I could work out how to link to these studies - I am a bit of a computer luddite (I'm learning!) but they are easy enough to find with a quick search.
As for the Swedish PM saying they got it wrong - Stefan Lofven is quoted on 22 August as saying Sweden got it right not to impose a strict lockdown.