Hmm, I'm not an expert statistician but I believe there's no chance of death if uninfected. Therefore, in the vaccinated group, that 1.02% chance of dying from Delta is part of that 0.42% chance of catching Delta in the first place.

Maybe other confounding factors are at play for why, once infected with Delta, vaccinated people may have a higher risk of death than the unvaccinated, such as higher rates of medical comorbidities or older people in the vaccinated group who caught Delta. But this is just my suspicion for now.

About the absolute risk reduction, I believe other experts have clarified that relative risk reduction is important too: https://www.reuters.com/article/factcheck-thelancet-riskreduction-idUSL2N2NK1XA

Because absolute risk reduction depends on the number of events, which are further influenced by other factors such as time period and population immunity. Who knows, maybe that 0.0001% survival would be higher as time passes, leading to more cases, or compared to pre-vaccine era with no population immunity.