The effect of teaching the immune system to attack something is difficult to predict because it is a natural system and we don't entirely understand how it works, only enough to get better at teaching it new things. In some cases it's so good at destroying related things that it produces "sterilising immunity" - a vaccinated person can't get the disease at all, even if directly exposed to it, in other cases, as seems to have been the case for COVID-19, you get something useful but much less effective.
AIUI There's no reason we know of why the platform (the mechanism by which we tell people's immune system about a new thing to be attacked) would influence how effective the defence put up subsequently is. The vaccine is not in your body when you're attacked weeks or months later by a virus, what's left is a "memory" of what should be identified as foreign and destroyed. So, it would be expected to depend on how AIDS works, and perhaps on exactly what the vaccine is teaching the immune system to attack, not on how the vaccine is manufactured.
Think about the difference between identifying wanted bank robbers with a hand-drawn picture of their faces, versus a still from a CCTV camera. It might be harder or easier to recognise the CCTV still, but we wouldn't expect them to do more or less jail time when identified by one method versus the other.
The data is showing that the likelihood of severe illness in a delta or omicron-infected person who previously had both doses of vaccine is significantly less.
...many more like that if you search.