It always is
The problem is the same as it's always been, models aren't accurate enough, in Norway we didn't expect the amount of cases of covid-19 related to Norwegians going skiing abroad (majority related to a bartender, obviously, as afterski is the most important part of going skiing), which meant that it spread faster than anticipated. Considering the amount of countries that have underestimated the rate of spread for their models, it'd hardly be shocking if the same goes for Sweden as well.
The argument that can be made is that if minor safety measures were introduced earlier, like telling people to reconsider going abroad for winter holiday etc, we'd achieve a decent enough rate to achieve herd immunity without the downside of closing shops and going into lockdown.
The plan in Norway is still herd immunity, we're just calling it something else for obvious reasons.