There are many factors here. Deaths are more likely among older people - we know that already. Deaths within 28 days, seem to be skewed by a high number of older people who only had the vaccine two weeks earlier - suggesting they were already infected and it didn't have time to work.

Other factors include the fact that it's only 90% effective at best against severe disease (10% of vulnerable people will still get very ill). That assumes it had 2-3 weeks to generate antibodies before people got infected.

Why is the number for unvaccinated lower? Probably because the number of unvaccinated people is much lower.

It's absolute numbers - not percentages of vaccinated vs percentages of unvaccinated, so the whole thing probably needs more statistical analysis to draw any real conclusions. I expect reading the full report would reveal more.