Vaccinated and unvaccinated individuals have similar viral loads in communities with a high prevalence of the SARS-CoV-2 delta variant: https://www.medrxiv.org/content/10.1101/2021.07.31.21261387v...
No Significant Difference in Viral Load Between Vaccinated and Unvaccinated, Asymptomatic and Symptomatic Groups Infected with SARS-CoV-2 Delta Variant: https://www.medrxiv.org/content/10.1101/2021.09.28.21264262v...
No analogy is perfect, but I think that this one is instructive, as it breaks the binary logic.
The statement "vaccines don't prevent..." conflates two very different statements, i.e. "vaccines don't prevent all occurrence of $outcome" (which is often true but not the point) and "vaccines don't have any impact on $outcome, prevent any occurrences of $outcome" (which is false, but is often implied).
It's clear that COVID-19-Omicron is infectious to a crazy degree, but that's a different statement to "this defensive measure does nothing".
Vaccines do not entirely prevent spreading. But nothing does.
Vaccines are not entirely useless to prevent spreading. It's one of the few measures that has an effect. Less so with Omicron, to be sure.
Would you say "When wearing a bulletproof vest, you can still be shot in the arm or head, so bulletproof vests do not prevent injury" (and might as well not wear one when in a dangerous situation)
Medicine is not Boolean logic, none of this is binary, all or nothing. We need risk reduction, not sweeping reductive statements.